Citation
Predicting health-related quality of life and benefit finding in Latina breast cancer survivor

Material Information

Title:
Predicting health-related quality of life and benefit finding in Latina breast cancer survivor
Creator:
Languido, Lauren
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Psychology, CU Denver
Degree Disciplines:
Clinical health psychology
Committee Chair:
Kilbourn, Kristin
Committee Members:
Risendal, Betsy
Borrayo, Evelinn
Ranby, Krista

Notes

Abstract:
As the Latino population in the United States (U.S.) continues to grow, so do health disparities, especially for those facing such serious illness as cancer. About 91.9 of every 100,000 Latinas were diagnosed with breast cancer between 2008 and 2012. Early detection and treatments have improved and have led to increased survivorship among Latinas. Survivorship involves unique challenges that may compound the challenges that many Latinas already face on a daily basis (e.g., socioeconomic challenges, limited or no health insurance, limited access to healthcare). Research has begun to evaluate how cultural and psychosocial factors influence health outcomes among Latina breast cancer survivors (LBCS). Data for the proposed study was derived from a sample of LBCS (n = 105) who participated in a multi-site, longitudinal, followup study, entitled “Survivorship Update Network to Southwest Hormone, Insulin, Nutrition, and Exercise [SUNSHINE],” which took place in Arizona and Colorado between April 2007 and July 2008. The study aimed to extend previous literature and examine potential predictors of health-related quality of life (HRQOL) and benefit finding (BF) in LBCS. Study findings showed that spiritual wellbeing independently predicted HRQOL and BF, and BF independently predicted HRQOL over and above age, marital status, education level, tumor stage and diagnosis, and comorbidities. However, perceived tangible social support did not predict HRQOL or BF in our sample. It is hoped that study findings will help improve understanding of how cultural factors might influence health outcomes among LBCS, as well as how intervention programs can be tailored to this demographically-defined group.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Lauren Languido. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
PREDICTING HEALTH-RELATED QUALITY OF LIFE AND BENEFIT FINDING IN LATINA
BREAST CANCER SURVIVORS By
LAUREN LANGUIDO
Post-Baccalaureate Certificate, Northwestern University, 2013 B.A., DePaul University, 2009
A thesis submitted to the Faculty of the Graduate School of the University of Colorado Denver in partial fulfillment of the requirements for the degree of Master of Arts
Clinical Health Psychology Program
2017


©2017
LAUREN LANGUIDO ALL RIGHTS RESERVED
11


This thesis for the Master of Arts degree by
Lauren Languido has been approved for the Clinical Health Psychology Program by
Kristin Kilbourn, Chair Betsy Risendal Evelinn Borrayo Krista Ranby
Date: December 16, 2017


Lauren Languido (B.A.)
Predicting Health-Related Quality of Life and Benefit Finding in Latina Breast Cancer Survivors Thesis directed by: Assistant Professor Kristin Kilboum
ABSTRACT
As the Latino population in the United States (U.S.) continues to grow, so do health disparities, especially for those facing such serious illness as cancer. About 91.9 of every 100,000 Latinas were diagnosed with breast cancer between 2008 and 2012. Early detection and treatments have improved and have led to increased survivorship among Latinas. Survivorship involves unique challenges that may compound the challenges that many Latinas already face on a daily basis (e.g., socioeconomic challenges, limited or no health insurance, limited access to healthcare). Research has begun to evaluate how cultural and psychosocial factors influence health outcomes among Latina breast cancer survivors (LBCS). Data for the proposed study was derived from a sample of LBCS (n = 105) who participated in a multi-site, longitudinal, followup study, entitled “Survivorship Update Network to Southwest Hormone, Insulin, Nutrition, and Exercise [SUNSHINE],” which took place in Arizona and Colorado between April 2007 and July 2008. The study aimed to extend previous literature and examine potential predictors of health-related quality of life (HRQOL) and benefit finding (BF) in LBCS. Study findings showed that spiritual wellbeing independently predicted HRQOL and BF, and BF independently predicted HRQOL over and above age, marital status, education level, tumor stage and diagnosis, and comorbidities. However, perceived tangible social support did not predict HRQOL or BF in our sample. It is hoped that study findings will help improve understanding of how cultural factors might influence health outcomes among LBCS, as well as how intervention programs can be tailored to this demographically-defined group.
IV


The form and content of this abstract are approved. I recommend its publication.
Approved: Kristin Kilboum
v


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION..............................................................1-13
Overview....................................................................1
Breast cancer incidence and prevalence among Latinas......................1-2
Health disparities and LBCS...............................................2-3
Theoretical Context.......................................................3-6
Health disparities and LBCS...............................................6-8
Perceived social support Among LBCS.......................................8-9
Benefit Binding and Cultural Lactors Among LBCS..........................9-11
The Present Study.......................................................11-13
II. METHODS..................................................................14-22
Sample.....................................................................14
Procedure..................................................................15
Measures................................................................15-21
Statistical Analysis....................................................21-22
III. RESULTS..................................................................23-33
Demographics and Descriptive Statistics in the Whole Sample.............23-25
Demographics and Descriptive Statistics in the LBCS Group...............25-26
Demographics and Descriptive Statistics in the NLWBCS Group................26
Assumptions.............................................................26-27
Bivariate Correlation Analyses with the Whole Sample....................27-28
Bivariate Correlation Analyses with the LBCS Group......................28-29
vi


Group Differences on Covariates......................................29-30
Group Differences on Primary Variables (Aim 1)..........................30
Hierarchical Multiple Regression Analyses (Aims 2-3) with the LBCS Group.30-33
IV. DISCUSSION............................................................34-41
V. REFERENCES............................................................42-51
vii


LIST OF TABLES
TABLE
1. Demographic Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), &
NLWBCS Group (n = 255).............................................................25-7
2. Descriptive Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), & NLWBCS
Group (n = 255)).....................................................................27
3. Bivariate Correlations of Covariates and Primary Variables in the Whole Sample (N =
360).................................................................................30
4. Bivariate Correlations of Covariates and Primary Variables in the LBCS Group (n = 105). ..31
5. One-Way Analysis of Variance (ANOVA) of Racial/Ethnic Differences on the Continuous
Covariate Age).......................................................................32
6. Tests of Racial/Ethnic Differences on Categorical Covariates.....................32
7. One-Way Analysis of Covariance (ANCOVA) of Racial/Ethnic Differences on Primary
Variables when controlling for Sociodemographic and Medical Covariates...............33
8. Results of Hierarchical Multiple Regression for SWB as a Predictor of HRQOL when
controlling for Sociodemographic and Medical Covariates in LBCS (n = 105)...........33
9. Results of Hierarchical Multiple Regression for SWB as a Predictor of BF when controlling
for Sociodemographic and Medical Covariates in LBCS (n = 105).......................34
10. Results of Hierarchical Multiple Regression for PTSS as a Predictor of HRQOL when
controlling for Sociodemographic and Medical Covariates in LBCS (n = 105)...........34-5
11. Results of Hierarchical Multiple Regression for PTSS as a Predictor of BF when controlling
for Sociodemographic and Medical Covariates in LBCS (n = 105).......................35
12. Results of Hierarchical Multiple Regression for BF as a Predictor of HRQOL when
controlling for Sociodemographic and Medical Covariates in LBCS (n = 105)...........35-6
viii


CHAPTER I
INTRODUCTION
Overview
Latinos make up the largest and fastest growing ethnic minority population in the United States (U.S.; Braveman et al., 2010; Graves et al., 2012). There are currently over 55 million Latinos living in the U.S., a number that is projected to reach 119 million by 2060 (U.S. Census Bureau, 2015). As the Latino population continues to grow, significant socioeconomic challenges and health issues that affect Latinos also increase (Ortega, Rodriguez, & Vargas Bustamante, 2015). In general, Latinos experience higher rates of poverty and adverse health outcomes, and often lack access to health care (Askim-Lovseth & Aldana, 2010; Gallo et al., 2009; Graves et al., 2012; Vega, Rodriguez, & Gruskin, 2009). Latinos often live without proper treatment for health issues until their illnesses have progressed. This is often the case for Latinos with cancer, especially Latina breast cancer patients. Latinas are often diagnosed at later stages of breast cancer and late-stage treatment typically involves more aggressive and costly interventions and poorer prognoses (Askim-Lovseth & Aldana, 2010; Ortega et al., 2015). Latina women with breast cancer face particularly significant health disparities compared to their Non-Latino White (NLW) counterparts (Yanez et al., 2011).
Breast cancer incidence and prevalence among Latinas
Although breast cancer incidence rates are lower among Latina women, breast cancer is the most common cancer and the most common cause of cancer-related deaths among Latinas. (American Cancer Society, 2015; Center for Disease Control and Prevention, 2015; Siegel et al., 2015). Between 2008 and 2012, approximately 91.9 of every 100,000 Latinas were diagnosed with breast cancer, and about 14.5 of every 100 Latinas died of breast cancer (Siegel et al.,
1


2015). More recently, Latinas have had greater exposure to better screening practices and access to improved treatment programs. Improved care has likely led to higher breast cancer survival rates and longevity among Latina breast cancer survivors (LBCS) both in general and in comparison to Non-Latina White breast cancer survivors (NLWBCS; Graves et al., 2012). However, LBCS in general continue to face significant challenges to their overall health, including lower education and socioeconomic status, lack of health insurance, limited access to healthcare, and poorer quality of life (QOL) compared to NLWBCS (Ruiz, Campos & Garcia, 2016; Rnheiro et al., 2011; Yanez et al., 2011).
Health disparities and LBCS
As mentioned above, research has shown that LBCS experience poorer overall health, quality of life (QOL), and health-related quality of life (HRQOL) compared to NLWBCS (Garcia-Jimenez et al., 2014; Graves et al., 2012; Martinez-Ramos et al., 2013; Yanez et al., 2011). In general, QOL is a multidimensional concept that reflects an individual’s sense of wellbeing derived from his/her satisfaction or dissatisfaction across a variety of life domains. Such life domains include physical, functional, psychological, spiritual, and/or social wellbeing. QOL is a broad term that may involve distinct cultural perspectives on wellbeing and may account for unique cultural contexts and value systems that may vary across cultural and ethnic groups (Kagawa-Singer, Padilla & Ashing-Giwa, 2010; Sammarco & Konecny, 2008). HRQOL is similar to QOL in that HRQOL involves the same range of wellbeing domains and may be defined and expressed differently across cultural and ethnic groups. However, HRQOL differs from general QOL in that it elucidates the individual’s perception of how his/her health and disease status impacts his/her overall wellbeing (Kagawa-Singer et al., 2010).
2


Ashing-Giwa et al. (2007) and Maly et al. (2008) conducted studies that demonstrated that LBCS experience poorer HRQOL compared to their African American and NLW counterparts. In a systematic review, Yanez et al. (2011) evaluated 22 studies comparing QOL among LBCS compared with other ethnic/racial groups. LBCS, in general across the studies, tended to report significantly poorer QOL compared to NLWBCS and experienced the largest differences across physical health, social wellbeing, and sexual health domains. In addition, a qualitative study revealed that LBCS primarily struggle with psychological concerns (e.g., crying and feeling sad, anxious, and irritable), followed by social functioning and spiritual and existential concerns.
In another qualitative study examining QOL among LBCS alone and in comparison to other ethnic groups, LBCS participants discussed the harmful side effects of treatment whereas NLWBCS discussed positive aspects of the breast cancer experience. Some quantitative studies confirmed findings that LBCS experienced poorer psychological and physical health compared to their NLWBCS counterparts. LBCS additionally reported more disruptions to social life and relationships than NLWBCS. Differences between LBCS and NLWBCS on outcome measures of QOL as well as psychological and physical health persisted even when researchers accounted for differences in treatment type, socioeconomic status (SES), and other sociodemographic variables. However, in the same vein, other studies found that LBCS experienced higher QOL and, as mentioned above, lower mortality rates and increased longevity compared to NLWBCS (Yanez et al., 2011).
Theoretical Framework
HRQOL and Contextual!Cultural Factors
3


Differences between LBCS and NLWBCS on health outcomes such as QOL may be due to differences in situational, contextual, and cultural factors (Garcia-Jimenez, 2013; Kawaga-Singer et al., 2010). Ashing-Giwa’s model of HRQOL highlights culture as a macro-component of HRQOL. Ashing-Giwa’s model of HRQOL might be better understood through the lens of Bronfenbrenner’s ecological model, especially because Ashing-Giwa employs a complex, contextual framework to explain HRQOL. This contextual framework involves the interrelationships among macro-level components (e.g., cultural, socio-ecological, demographic, health care systems factors, etc.) and micro- or individual level components (e.g., health status, disease characteristics, health efficacy, level of functioning, etc.) (Bronfenbrenner, 1994; Kagawa-Singer et al., 2010; Ashing-Giwa et al., 2007). The model’s contextual framework also involves the unique and interacting influences of macro- and micro-level components on the various dimensions of wellbeing. Ashing-Giwa’s model of HRQOL included the same dimensions of HRQOL as other measures of HRQOL and general QOL. Ashing-Giwa’s model of HRQOL focused on the following dimensions of HRQOL: physical wellbeing, functional wellbeing, psychological and emotional wellbeing, social wellbeing, spiritual wellbeing, and sexual wellbeing (Kagawa-Singer et al., 2010; Ashing-Giwa et al., 2007).
LBCS in the U.S. generally face high adversity such as lower income, education, and employment opportunities, as well as higher rates of discrimination, and limited or lack of access to health insurance and quality care (Ruiz et al., 2016). However, despite high adversity, LBCS tend to experience lower mortality rates and increased longevity compared to NLWBCS. This comparative advantage is an important example of the Latino Health Paradox (Pinheiro et al., 2011; Viruell-Fuentes & Schulz, 2009). The Latino Health Paradox is an epidemiological phenomenon in which Latinos, in general, overcome adversity and experience more health
4


advantages compared to Non-Latino Whites with regard to physical and psychological health outcomes. Recently, research has shown that Latinos’ comparative advantages may be due to cultural characteristics such as strong social integration and cohesion, and a strong emphasis on spirituality and/or religiosity (Gallo et al., 2009; Patel et al., 2013; Ruiz et al., 2016).
The Latino Health Paradox and Cultural Values
Some research suggests that adherence to some Latino cultural values might lead to better health outcomes (Gallo et al., 2009; Patel et al., 2013; Ruiz et al., 2016). For example, more ethnically dense Latino neighborhoods are associated with better physical and psychological health for both Latinos and Non-Latinos (Ruiz et al., 2016). Latino cultural values related to interpersonal relationships are commonly cited in the literature. Such values include familismo (focus on family), personalismo (personal versus institutional relationships), respeto (mutual and reciprocal respect), confianza (trust and intimacy in relationships), collectivism, and family!social support (Anez et al., 2005; Gallo et al., 2009). Similarly, religiosity and/or spirituality are Latino cultural values that are often viewed and experienced as enmeshed concepts. Religiosity and/or spirituality are often understood as relating to the interpersonal relationships that one creates with a religious and/or spiritual community as well as the intrapersonal experience of relating to a higher power (Hunter-Hernandez et al., 2015). Religiosity and spirituality as well as social support have been especially highlighted in research with LBCS as protective cultural factors. Such factors have also been found to reduce risk and improve both resilience and health outcomes among Latinos in general (Hunter-Hernandez et al., 2015; Gallo et al., 2009).
However, Latinos might experience reduced resilience and higher risk of adverse health outcomes when they lose or lack protective Latino cultural factors. One such Latino cultural factor is social support. LBCS who do not have access to social support may experience a higher
5


risk of recurrence and comorbid physical and psychological health issues in the survivorship period (Schwartz et al., 2013; Rasheed et al., 2011; Napoles et al., 2011). A history of cancer has been found to be associated with breast-cancer specific mortality significantly more often in Latinas than any other ethnic/racial group (Wu et al., 2014). In general, comorbidities such as diabetes and obesity have been found to be associated with breast-cancer specific mortality regardless of ethnic/racial group (Connor et al., 2016; Wu et al., 2014). Connor et al (2016) found that LBCS were more likely to have a history of diabetes and to be classified as obese compared to their NLW counterparts, which may partially explain why comorbid diabetes and obesity are more highly associated with breast-cancer specific mortality in LBCS.
In addition, Ashing et al. (2014) assessed comorbidities in breast cancer survivorship among LBCS (n = 232) and African American BCS (// = 88) given that both groups tend to face significant health disparities. The researchers found that LBCS reported twice as many headaches and migraines, as well as significantly more osteoporosis than African American BCS (Ashing et al., 2014). Given the foregoing weight of the evidence in the literature, the factors that contribute to health outcomes and overall QOL in LBCS are not well understood. Asa result, Latina breast cancer survivorship and the mechanisms contributing to risk and resilience are of prime interest to clinical researchers (Martinez-Ramos, Biggs & Lozano, 2013). More specifically, research is needed to further explore the relationships between specific cultural values and health outcomes (Ruiz et al., 2016).
Spiritual Wellbeing among LBCS
Spiritual wellbeing (SWB) is conceptualized as a multidimensional construct that may encompass both religious and non-religious derivations of faith and inner peace (Garcfa-Jimenez et al., 2014). Spirituality generally refers to an individual’s internal experience of faith, hope, and
6


meaning, whereas religiosity refers to the behaviors and rituals in which an individual engages and that are based on organized religion. However, as mentioned above, many individuals may experience spirituality and religiosity as enmeshed concepts (Levine, Yoo, Aviv, Ewing & Au, 2007). For many Latinos, spirituality is considered intrinsic to Latino culture (Yanez et al., 2011; Gallo et al., 2009). Latinos in general may adhere to a broad range of spiritual and religious beliefs, values, and practices that may vary according to gender, geographic region, specific Latino sub-culture (e.g., Mexican, Guatemalan, Cuban, Ecuadorian, Spanish, etc.) (Hodge et al., 2013). Regardless of such differences, Latinos as a group in the U.S. are recognized for their high rates of spirituality and religiosity compared to NLW. For example, about 55 percent of the 35.4 million Latino adults in the U.S. consider themselves Catholic, followed by other Christian religions, and then by non-Christian religions (Pew Research Center, 2008).
Unlike NLW in the U.S. who tend to experience spirituality and religion as separate from daily life, Latino individuals and communities experience spirituality and religion as infused in daily life. In this way, Latino culture embraces and influences spiritual and religious interpretation and expression among Latinos (Hodge at al., 2013). Latinos are known to engage in spiritual and religious coping when dealing with a variety of difficult life situations including cancer diagnosis, treatment, and survivorship (Yanez et al., 2011). Gallo et al. (2009) highlighted several studies that demonstrated that higher levels of spiritual and religious coping were associated with higher levels of self-rated satisfaction with life, a reduction of harmful health behaviors, and higher treatment-seeking among substance users of diverse populations including Latinos (Gallo et al., 2009). With regard to LBCS, Levine et al. (2007) used a qualitative approach including open-ended questions to evaluate coping styles among diverse BCS. Based
7


on this approach, the researchers found that LBCS reported spirituality and/or religiosity as key coping strategies.
Research has further highlighted the importance of SWB for oncology patients, including breast cancer patients and survivors, and especially for Latino cancer patients and survivors. Foundational research evidences SWB as a significant predictor of HRQOL among LBCS (Yanez et al., 2011). Brady et al. (1999) found that SWB has unique effects on both subjective and objective health outcomes and HRQOL among people who have had a serious illness(es) including breast cancer. Bredle et al., (2011) confirmed these findings. In addition, individuals who engage in organized religious activities tend to experience better psychological and physical health outcomes (Gallo et al., 2009). Moreover, Yanez et al. (2011) highlighted multiple studies that associated higher SWB with higher reports of HRQOL, and demonstrated that Latinas with breast cancer utilized spiritual and/or religious coping more frequently than NLW with breast cancer. As a result, SWB, similar to social support, may serve as a protective factor against adverse health outcomes for Latina breast cancer patients and survivors (Yanez et al. 2011). Perceived social support Among LBCS
Social support is an important Latino cultural value. Social support may influence health outcomes depending on one’s perception of how much social support he/she experiences. In general, perceived social support involves an individual’s beliefs and experiences about how much he/she is cared for, esteemed, and valued, as well as the extent to which one perceives him-/herself to be connected to a social network (e.g., family, friends, community). One’s social network might also involve people to whom one can turn for emotional and/or tangible (e.g., material, financial, etc.) social support (Crookes et al., 2016). There is a considerable body of research on the relationship between social support and QOL among cancer patients and
8


survivors in general. Kroenke et al. (2013b) found that positive social interaction was significantly related to multiple measures of QOL including breast cancer-specific QOL, HRQOL, physical wellbeing, social wellbeing, and emotional wellbeing. In previous work, Kroenke et al. (2013a; 2012) found that the association between social networks and breast cancer outcomes was related to high and low levels of social support (i.e., emotional support, tangible support, affection, and positive interaction) and social burden (i.e., caregiving responsibilities, social strain or negative aspects of social relationships) within relationships. For example, low levels of support and high relationship burden were independently associated with higher mortality rates among women with cancer.
Among LBCS, the perception of social support can influence QOL outcomes. In Yanez et al.’s (2011) systematic review, multiple studies found that perceived social support in general was positively correlated with QOL among LBCS. Graves et al. (2012) similarly found that perceived social support was significantly and positively associated with QOL. Additionally, among immigrant LBCS, those who reported less perceived social support tended to also report poorer QOL compared to other racial/ethnic groups. Immigrant LBCS often leave behind family and friends when they immigrate to the U.S. and may take time to build new social connections (Crookes et al., 2016). However, social support is not only an important predictor of QOL but it has been found to be an important predictor of how much benefit cancer survivors may glean from the cancer experience (Weaver at al., 2008).
Benefit Finding and Cultural Factors Among LBCS
There is a substantial and growing body of research on the positive ways in which one’s life may change after a traumatic event or illness such as cancer. Researchers often label these positive changes as “post-traumatic growth”, “stress-related growth”, or “benefit finding” (BF)
9


(Hegelson, Reynolds & Tomich, 2006). Although theories on BF vary slightly, they all agree that BF involves the ability to find benefits and/or make positive life changes following a traumatic event (e.g., cancer diagnosis, treatment, etc.). BF may further be conceptualized as a coping strategy by which cancer patients and survivors may manage acute and chronic health issues and find meaning in the experience. Finding meaning in one’s experience may be especially useful for those who face the long-term effects of their cancer and treatment such that they may regain a perceived sense of mastery and control over their situation. In these ways, BF can influence physical and psychological health outcomes (Pascoe & Edvardsson, 2013).
Previous research with BCS showed that higher perceived benefits were generally associated with better coping, higher QOL, and more positive overall health outcomes (Weaver et al., 2008; Tomich & Hegelson, 2004). Hegelson, Reynolds & Tomich (2006) conducted a meta-analysis of 87 studies and found that BF was associated with less depression and greater positive wellbeing across diverse racial/ethnic populations who experienced a variety of traumatic experiences (e.g., cancer, war-stress, etc.). Tomich & Hegelson (2004) found that minority (i.e., Latinas and African American) women with breast cancer tended to perceive greater benefits, (t(36Q) = -3.01, p = .01) and to have less negative affect (1(360) = 4.09, p = .01) compared to NLW with breast cancer. Researchers suggested that lower SES and greater disease severity led minority women to engage more in BF compared to NLW women. Researchers also indicated that relationships among low SES, race/ethnicity, and BF might have been due to spiritual and/or religious coping strategies that involve cognitive restructuring of events. Additionally, individuals who face high adversity in general may be more apt to find benefit in negative events. However, Tomich & Hegelson (2004) also found that greater BF, regardless of race/ethnicity, was marginally associated with negative affect (r = .09, p = .10).
10


Given all this, research findings suggest that BF might be both an important coping strategy and psychological health outcome for such minority groups as LBCS. However, more research is needed to elucidate which factors or mechanisms predict BF and HRQOL among LBCS (Pascoe & Edvardsson, 2013; Garcia-Jimenez, 2013; Graves et al., 2012; Martinez-Ramos et al., 2013; Yanez et al., 2011). In particular, a better understanding of how spirituality and social support may function as protective factors to improve BF and HRQOL in LBCS is necessary. Such data can then be used to highlight and explain the risks that LBCS face, as well as new ways in which cancer survivorship programs may be improved for growing Latino communities in the U.S. In this way, future generations of Latinos may face fewer challenges and threats to overall physical and psychological health.
The Present Study
The purpose of this study was to explore the relationships among spiritual wellbeing, perceived tangible social support, health-related quality of life, and benefit finding in a sample of Latina breast cancer survivors (LBCS) from Arizona and Colorado. The follow-up study SUNSHINE is a population-based, case-control study that has allowed researchers to evaluate the differences between LBCS and Non-Latino White breast cancer survivors (NLWBCS) with regard to breast cancer risk based on genetic and lifestyle factors. The sample population was also at least 6 years post-treatment, such that they may be considered long-term cancer survivors. According to the American Cancer Society, a long-term cancer survivor is at least 5 years posttreatment (Chopra & Kamal, 2012). SUNSHINE utilized reliable markers of sociodemographic variables (e.g., age, marital status), medical variables (e.g., tumor stage at diagnosis, comorbidities), as well as such psychosocial variables as quality of life, spiritual wellbeing, perceived tangible social support, and benefit finding (Risendal et al., 2014). Therefore,
11


SUNSHINE offers an excellent opportunity to explore the relationships among such psychosocial variables among LBCS while controlling for differences on confounding variables.
A priori power analyses were conducted to determine whether or not the proposed sample sizes from the SUNSHINE project would be large enough to detect a medium effect size using the proposed analyses. The first power analysis revealed that a total sample size of at least 128 is required to detect a medium effect size (f= .25) using the proposed analytic approach to test the preliminary hypotheses under Aim 1. Thus, the total sample of BCS cases (N = 360) will be sufficient to achieve enough statistical power to detect a medium effect size. The second power analysis revealed that a total sample size of 55 LBCS cases would be required to achieve enough statistical power to detect a medium effect size (f= .15) using the proposed analytic approach to test the proposed hypotheses under Aims 2-4. Thus, the total sub-sample of LBCS cases (n =
105) was proposed to be sufficient to achieve enough statistical power to detect a medium effect size.
Hypotheses
To evaluate the independent effects of perceived spiritual wellbeing and perceived tangible social support on health-related quality of life and benefit finding in the survivorship period for LBCS, the following aims and hypotheses will be evaluated:
Aim 1) To examine whether there are differences between LBCS and NLWBCS on measures of spiritual wellbeing, perceived tangible social support, HRQOL, and BE when controlling for sociodemographic (i.e., age, education level, marital status) and medical variables (i.e., tumor stage at diagnosis, comorbidities).
Hypothesis la: LBCS will report higher spiritual wellbeing than NLWBCS when controlling for sociodemographic and medical variables.
12


Hypothesis lb: LBCS will report higher perceived tangible social support than NLWBCS when controlling for sociodemographic and medical variables.
Hypothesis 1c: LBCS will report lower HRQOL than NLWBCS when controlling for sociodemographic and medical variables.
Hypothesis Id: LBCS will report higher benefit finding than NLWBCS when controlling for sociodemographic and medical variables.
Aim 2) To examine whether spiritual wellbeing is a significant predictor of HRQOL and benefit finding in LBCS.
Hypothesis 2a: Spiritual wellbeing will be a significant predictor of HRQOL in LBCS when controlling for sociodemographic and medical variables, such that higher levels of self-reported spiritual wellbeing predicts better HRQOL.
Hypothesis 2b: Spiritual wellbeing will be a significant predictor of benefit finding in LBCS when controlling for sociodemographic and medical variables, such that higher levels of spiritual wellbeing predict more benefit finding.
Aim 3) To examine whether perceived tangible social support is a significant predictor of HRQOL and benefit finding in LBCS.
Hypothesis 3a: Perceived tangible social support will be a significant predictor of HRQOL in LBCS when controlling for sociodemographic and medical variables, such that higher levels of perceived tangible social support predict better HRQOL.
Hypothesis 3b: Perceived tangible social support will be a significant predictor of benefit finding in LBCS when controlling for sociodemographic and medical variables, such that higher levels of perceived tangible social support predict more benefit finding. Exploratory Aim 4) Examine whether benefit finding predicts HRQOL in LBCS.
13


Hypothesis 4: Benefit finding will predict HRQOL in LBCS when controlling for sociodemographic and medical variables such that more benefit finding predicts better HRQOL.
14


Chapter II
Methods
Sample
The Survivorship Update Network to Southwest Hormone, Insulin, Nutrition, and Exercise (SUNSHINE) (N = 725) is a population-based, case-control study that has allowed researchers to evaluate the differences between Latina breast cancer survivors (LBCS) and Non-Latina breast cancer survivors (NLWBCS) from Arizona and Colorado with regard to breast cancer risk based on lifestyle and genetic factors (Risendal et al., 2014). SUNSHINE took place in Arizona and Colorado between April 2007 and July 2008. Although 1,969 women completed the parent study [i.e., Southwest Hormone, Insulin, Nutrition, and Exercise (SHINE) 4-Corners Breast Cancer Study], those who met the following criteria were considered ineligible for the follow-up study SUNSHINE: 1) they were deceased; 2) they had a mental or physical disability that would inhibit them from completing the interview; 3) they were living in a nursing home or group facility. There were 1,969 women who completed the study. Of the 1,969 women who completed the SHINE study, about 1,572 women were considered eligible for SUNSHINE. About 1,068 consented to participate in SUNSHINE. However, data was considered incomplete if participants completed either the written interview only (n = 70) or the telephone interview only (n = 150). Data was also considered incomplete if participants provided incomplete surveys (n = 123). Given all this, about 725 participants were found to have provided complete data.
Because the present study aimed to evaluate the differences between LBCS and NLWBCS, all control studies (N = 365) from SUNSHINE were excluded. Case studies (// = 360) were categorized into LBCS (n = 105). Additionally, women were excluded from the analyses if they reported a recurrence and/or more than 1 occurrence of breast cancer. In these ways, the
15


analyses might better shed light on the differences between LBCS and NLWBCS on variables known to be important in BCS populations (Aim 1). The analyses might also shed light on variables that may serve as key predictors of HRQOL and BF in LBCS (Aims 2-4).
Procedure
Research assistants at the Research Core, University of Colorado Cancer Center (Aurora, CO), screened and recruited participants from the SHINE 4-Corners Breast Cancer Study. Descriptions of study design and recruitment methods of this parent study have been published (Risendal et al., 2015; Sedjo et al., 2013). Between April 2007 and July 2008, research assistants randomly selected and contacted 20-40 potential participants each week. The study employed the Dillman protocol for recruitment, which involved the following steps: 1) Research assistants mailed an introductory letter and project brochure; 2) Two weeks later, research assistants mailed a packet that contained the introductory letter, project brochure, and a written questionnaire; 3) If participants did not respond within four to six weeks, research assistants sent reminder postcards to the participants (Risendal et al., 2014; Hoddnott & Bass, 1986). Research assistants then scheduled telephone interviews with consenting participants. Data that was not collected via written surveys was again collected via telephone interviews in addition to other questionnaires. The questionnaires allowed researchers to collect a variety of sociodemographic, medical, and psychosocial data (Risendal et al., 2014). As mentioned above, study analyses first focused on all breast cancer cases (N = 360) (Aim 1), and then focused solely on LBCS (// = 105) (Aims 2-4). Measures Spiritual wellbeing
Spiritual wellbeing was assessed with the Functional Assessment of Chronic Illness Therapy - Spiritual Wellbeing Scale version 4 (FACIT-Sp-12) and was available in English and
16


Spanish. The FACIT-Sp-12 is a unidimensional, quantitative measure that has been widely used and validated with cancer populations. The FACIT-Sp-12 is a 12-item self-report measure that assesses two aspects of spiritual wellbeing: Meaning/Peace and Faith. Items were meant to address a respondent’s sense of meaning in life, harmony, peacefulness, and to what extent the respondent is able to derive strength and comfort from his/her faith. The FACIT-Sp-12 utilizes 5-point Likert-style responses anchored as: (0) “Not at all(1) “A little bit(2) “Somewhat(3)
“Quite a bit(4) “Very much." Responses may be summed, yielding a score between 0 and 48 with higher scores indicating higher spiritual wellbeing. The FACIT-Sp-12 was found to have high internal consistency (Cronbach’s a = .87). The subscales were found to be moderately and positively correlated with the FACT-G (rs = .58, p < .001). However, the Meaning/Peace subscale was more strongly (positively) correlated with the FACT-G (rs = .62, p < .001). The FACIT-Sp-12 also demonstrated substantial discriminant validity such that there was an inverse relationship between FACIT-Sp-12 scores and POMS depression subscale scores [F(2, 1586) = 186.98, p= .0001],
Perceived tangible social support
Perceived tangible social support was assessed using the tangible social support subscale from the 40-item Interpersonal Support Evaluation List (ISEL). The perceived tangible social support subscale includes 10 items and has been used to assess one’s perception of the material or financial aid available to him/her, as well as support with daily chores or activities (Uchino, 2004) (e.g., “If I needed help fixing an appliance or repairing my car, there is someone who would help me,” “It would me difficult to find someone who would lend me their car for a few hours,” “If I had to go out of town for a few weeks, it would be difficult to find someone who would look after my house or apartment (the plants, pets, garden, etc.).” The ISEL and
17


subsequent versions (e.g., ISEL - Short Form, ISEL-12) have been widely accepted and utilized in psychosocial research, especially as a predictor of psychological and physical health indices (Uchino, 2004). The ISEL overall demonstrates good internal consistency (Cronbach’s a = .68). More specifically, the tangible social support subscale demonstrated good test-retest reliability (.69) within a student population. The perceived tangible social support subscale is available in English and Spanish. The tangible social support subscale utilizes 4-point Likert-style responses anchored as: (0) “Definitely true”; (1) “Probably True”; (2) “Probably false”; (3) “Definitely false” (Cohen & Hoberman, 1983).
Research has demonstrated the psychometric validity and reliability of the ISEL as a general population measure (Brookings & Bolton, 1988). Brookings & Bolton (1988) conducted confirmatory factor analysis (CFA) of the ISEL with 133 college students who participated in a longitudinal study of stress and depression. The CFA demonstrated that the ISEL has substantial structural validity such that there are 4 factors of perceived social support, including tangible, belonging, appraisal, and self-esteem (Brookings & Bolton, 1988). The ISEL was also found to have high convergent validity with other established measures of social support, as well as sufficient discriminant validity with personality measures (Cohen, Mermelstein, Kamarck, & Hoberman, 1985, as cited in Brookings & Bolton, 1988).
HROOL
HRQOL was assessed with the Functional Assessment of Cancer Therapy Scale (FACT-G) version 4 and was available in English and Spanish. The FACT-G version 4 is a 27-item self-report measure, in which items evaluate four dimensions of wellbeing: physical (PWB), functional (FWB), social/family (SFWB), and emotional (EWB). The PWB subscale refers to physical symptoms that the respondent may have experienced. The FWB refers to the extent to
18


which a respondent is able to participate in and enjoy everyday activities. The SFWB subscale refers to the types of social support one might receive from family and friends. The SFWB subscale includes 7 items that has been used to assess one’s perception of closeness or emotional support received from family and friends (e.g., “I feel close to my friends,” “I get emotional support from my family”). In this way, SFWB subscale addresses different aspects of social experience compared to the tangible social support subscale of the ISEL. The EWB subscale refers to the respondent’s general mood and emotional response to the illness (Ashing-Giwa & Rosales, 2013). All of the subscales utilize 5-point Likert-style responses anchored as: (0) “Not at all”; (1) “A little bit”; (2) “Somewhat”; (3) “Quite a bit”; and (4) “Very much.”To score the FACT-G, participants’ responses may be summed, yielding a score between 0 and 112. Higher scores indicate higher HRQOL (Celia et al., 1993).
The FACT-G is a uni dimensional, quantitative measure that is considered one of the most reliable and valid measures of HRQOL in chronic illness populations including cancer (Ashing-Giwa & Rosales, 2013). In terms of reliability, analyses revealed an overall Cronbach’s alpha score of .90 on the English version. The subscales had similarly high alpha coefficient scores (PWB, .82; FWB, .80; SFWB, .69; EWB, .74). In terms of validity, the FACT-G has shown substantial convergent validity such that it was significantly correlated (r = .79, p < .05) with another well-established QOL measure, the Functional Living Index-Cancer measure, as well as with the well-established mood distress scale Taylor Manifest Anxiety Scale (r = .58, p < .05) and Brief Profile of Mood States (POMS)(r = - 0.65, p < 0.05). The FACT-G has also shown substantial divergent validity such that it has been found to have a low correlation with the Marlowe-Crowne Social Desirability Scale (r = .22, p < .05). In addition, the FACT-G has been
19


found to be highly sensitive to disease stage, especially on the PWB and FWB subscales (Celia et al., 1993).
Moreover, the FACT-G Spanish version has been found to have high concurrent validity with the English version of the FACT-G, as well as high content and semantic validity (Celia et al., 1998). The FACT-G Spanish version has also been more specifically cross-validated with Latina breast cancer populations. The FACT-G demonstrated high construct validity such that the four factors accounted for 56 percent of the variance in HRQOL among English-language proficient Latinas and 57 percent among limited English-language proficient Latinas. The FACT-G also demonstrated high internal consistency (Cronbach’s a = .91) among English-language proficient and limited English-language proficient Latinas with breast cancer (Ashing-Giwa & Rosales, 2013).
Perceived benefit of the breast cancer experience
Perceived benefit or benefit finding (BF) was assessed using the Benefit Finding (BF) Scale for breast cancer, which was available in English and Spanish. The BF scale is a 17-item self-report measure that assesses the benefits that a respondent may perceive as a result of diagnosis and treatment of breast cancer. The BF scale is a uni dimensional, quantitative measure that has been commonly used and validated in cancer populations. Subscales include: acceptance, sensitivity to others, improved coping, and new purpose of life. The BF scale utilizes 5-point Likert-style responses anchored as: (0) “Not at all”; (1) “A little”; (2) “Moderately”; (3) “Quite a bit”; (4) “Extremely.”To score the BF scale, participants’ responses may be summed, yielding a score between 0 and 85. Higher scores indicate higher perceived benefit. The BF scale was found to have high internal consistency (Cronbach’s a = .95). The BF scale was also found to have convergent and discriminant validity. It was somewhat positively correlated with
20


optimism (r = .23), and inversely (but again not overwhelmingly) correlated with a POMS-derived index of distress (r = -.25) (Boyer et al., 2000 in Antoni et al., 2001).
Race/Ethnicity
Race/ethnicity was measured as a single item in order to observe differences between NLW and Latina breast cancer survivors. As mentioned above, previous literature (Ashing-Giwa et al., 2007; Maly et al., 2008; Yanez et al., 2011), demonstrates significant differences among racial/ethnic groups on survivorship outcomes, including NLW and Latina breast cancer survivors.
Covariates
Covariates were included in order to increase statistical power and reduce the probability of a Type II error. The covariates assessed in the present study included sociodemographic and medical variables. These covariates were chosen based on previous analyses from the same dataset (Risendal et al., 2015; Sedjo et al., 2013), as well as previous literature on the topic of breast cancer survivorship in multiethnic samples including LBCS, (Garcia-Jimenez et al., 2013; Janz et al., 2014; Maly et al., 2008), which suggest that the covariates chosen for this project might account for some of the variance in such outcomes as HRQOL and BL in breast cancer survivors, including LBCS.
Sociodemographic variables. The sociodemographic variables that were included in the analysis of covariance (ANCOVA) analyses were race/ethnicity, age, education level, and marital status. The sociodemographic variables that were included in the multiple regression models did not include race/ethnicity, and only included the following sociodemographic variables: age, education level, and marital status.
21


Age at the time participants were interviewed was measured as a single item and was coded as a continuous variable. Education level was measured as a single item and was coded as a categorical variable with 4 categories, which included the following: 1 = < High School; 2 = High School; 3 = Some college; and 4 = College +. Marital status was measured as a single item and was coded as a categorical variable with 6 categories, which included the following: 1 = Married; 2 = Living as married; 3 = Widowed; 4 = Divorced; 5 = Separated; and 6 = Never married. In order to account for the differences between having a long-term partner (defined as married or living as married) and being single, the variable was re-coded and entered into each multiple regression analysis as a binary, dummy variable as follows: 1 = Married or Living as married; and 2 = Currently single (widowed, divorced, separated, never married).
Medical variables. The medical variables that were included in the multiple regression models were tumor stage at diagnosis and the number of self-reported comorbidities. Tumor stage at diagnosis was measured using the Surveillance Epidemiology and End Result (SEER) cancer registry or the state tumor registry database. The second edition of the International Classification of Diseases for Oncology (ICD-O-2) was used for breast cancer cases diagnosed prior to 2001, whereas the ICD-O-3 was used to code cases diagnosed after 2001. The histologic types were categorized as follows: ductal carcinoma (8230, 8500, 8521, 8523), lobular carcinoma (8520, 8524), ductal/lobular (8522), all others, and unknown. Tumor stage classifications were then determined according to SEER summary stage codes using the 1977 definitions for breast cancer cases diagnosed prior to 2001. For breast cancer cases diagnosed after 2001, the 2000 definitions were used. The tumor stage at diagnosis variable was then coded as a categorical variable as follows: 0 = In situ (i.e., pathological stage 0); 1 = Local; 2 = Regional by direct extension (D.E.); 3 = Regional lymph nodes only involved; 4 = Regional by
22


both D.E. and to regional nodes; 5 = Regional not otherwise specified (NOS); 7 = Distant sites and/or distant nodes; 8 = CNS (benign or borderline); and 9 = Unknown or Not applicable. However, previous analyses of the dataset condensed the variables into the following: 0 = In situ; 1 = Localized; 2 = Regional/Distant; 3 = Unstaged or Missing (Sedjo et al., 2013). Therefore, for analyses in the present study, and to account for the differences between non-metastasized and metastasized cancer, the tumor stage at diagnosis variable was entered into the statistical models as a binary, dummy variable as follows: 1 = In situ or localized; and 2 = Regional or regional NOS or distant.
Comorbidities were measured using the Charlson Comorbidity Index (CCI). The CCI allows researchers to categorize participants’ comorbidities according to the International Classification of Diseases 9. The CCI provides a weight for each comorbidity based on the adjusted risk of mortality or the amount of healthcare resources utilized by a patient. The weights are summed to produce a single comorbidity score for each participant. Higher CCI scores are associated with an increase in the adjusted risk of mortality or higher resource use (University of Manitoba, 2016). The CCI variable used in the present study was measured categorically between 0 and 17 comorbidity categories. Because not many participants had 3 or more comorbidities, it was not useful to maintain 17 categories. Most participants were distributed across 3 levels of categories; therefore, this variable was entered into the statistical models as a tertile variable as follows: 0 = 0; 1 = 1 or 2; 3 = 3 to 14.
Statistical Analysis
Before conducting the main analysis, data screening was conducted and assumptions for ANCOVA and hierarchical multiple regression analyses were tested. In order to conduct such multivariate analyses, IBM SPSS Statistics for Windows, version 24 (IBM Corp., Armonk, N.Y.,
23


USA) was used. The significance level for all statistical analyses was a < .05 (two-tailed). Descriptive statistics of the Whole Sample (N = 360), LBCS group (// = 105), and the NLWBCS (n = 255), respectively, were subsequently computed, which included means, standard deviations, frequencies, and percentages (see Tables 1 and 2). Next, bivariate correlations between all pairs of variables of the Whole Sample (N = 360) and the LBCS group (// = 105), respectively, were conducted (see Tables 4 and 5). These bivariate correlations allowed for the interpretation of direction and strength of any potential correlations between each pair of variables.
Based on the bivariate correlation models between predictor and outcome variables, multicollinearity was not observed among the hypothesized covariates (i.e., age, education level, marital status, tumor stage at diagnosis, and comorbidities). However, the hypothesized covariates were correlated with the primary variables of interest; therefore, they were maintained across all analyses. ANCOVAs were then conducted to evaluate Aim 1 or group mean differences on the primary variables (i.e., SWB, PTSS, HRQOL, and BF). In order to evaluate Aims 2-4, hierarchical multiple linear regressions were conducted, because hierarchical multiple linear regression allows researchers to specify a fixed order of entry for variables, and as a result, control for the effects of specific covariates. The models generally included sociodemographic variables in the first block, medical variables in the second block, and the respective predictor variable in the third block. Overall, the statistical models illustrated the extent to which each predictor variable was independently associated with respective outcome variables above and beyond the effects of the sociodemographic and medical variables.
24


Chapter III
Results
Demographics and Descriptive Statistics in the Whole Sample
The total sample of breast cancer survivor cases (N = 360) included more NLWBCS (71%) than LBCS (29%). Participants’ ages ranged from 33 to 85 (M= 61.37; SI) = 9.99). Most participants completed at least some college (73%) and were married or living as married (74%). More participants were initially diagnosed with in situ or local cancer (68%) compared to regional, regional NOS, or distant cancer (26%). With regard to comorbidities, 52% of all breast cancer survivors reported a Charlson Comorbidity Index (CCI) score of 0, 27% had a CCI score of 1 or 2, and 19% had a CCI score of 3 - 14. Table 1 (below) provides an overview of the demographic information for the Whole Sample. Table 2 (below) provides an overview of the descriptive information for the Whole Sample.
Table 1. Demographic Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), &
NLWBCS Group (n = 255) Participant Characteristic Age Minimum Maximum M(SD) n %
Whole Sample 33 85 61.37 (9.99) 360
LBCS 33 83 59.19(10.64) 105
NLWBCS 43 85 62.27(9.59) 255
Race/Ethnicity
Latina 105 29.2
Non-Latina White 255 70.8
Education level
Whole Sample
< HS 34 9.4
HS 65 18.1
Some college 130 36.1
College+ 131 36.4
LBCS
HS 22 21.0
Some college 36 34.3
25


Table 1 cont’d
College+
NLWBCS
HS
Some college College+
Marital status Whole Sample
Married or living as married
Single, widowed, divorced LBCS
Married or living as married
Single, widowed, divorced NLWBCS
Married or living as Married
Single, widowed, divorced
Tumor stage at diagnosis Whole Sample
In situ or localized Regional, regional NOS, or distant cancer LBCS
In situ or localized Regional, regional NOS, or distant cancer NLWBCS
In situ or localized Regional, regional NOS, or distant cancer
CCI score
Whole Sample 0
20
7
43
94
111
268
92
74
31
194
61
246
92
68
29
178
63
195
19.0
2.7
16.9
36.9
43.5
74.4
25.6
70.5
29.5
76.1
23.9
68.3
25.6
64.8
27.6
69.8
24.7
54.2
26


Table 1 cont’d
1 or 2 96 26.7
3 - 14 69 19.2
LBCS
0 47 44.8
1 or 2 26 24.8
3 - 14 32 30.5
NLWBCS
0 148 58
1 or 2 70 27.5
3-14 37 14.5
Note. HS = high school; NOS = not otherwise specified; CCI = Charlson comorbidity index
Table 2. Descriptive Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), &
NLWBCS Group (n = 255)
Participant Characteristic Minimum Maximum M(SD) N
SWB
Whole Sample 5 48 37.96(8.36) 356
LBCS 13 48 38.84( 8.09) 104
NLWBCS 4 48 37.60(8.45) 252
PTSS
Whole Sample 10 38 23.94(3.18) 353
LBCS 10 38 23.45(4.13) 104
NLWBCS 10 37 24.15(2.67) 249
HRQOL
Whole Sample 34.40 112 92.74(14.80) 348
LBCS 34.40 112 88.59(16.70) 102
NLWBCS 49 112 94.46(13.62) 246
BF
Whole Sample 17 85 59.56(14.94) 342
LBCS 30 85 64.21(12.611) 101
NLWBCS 17 85 57.61(15.43) 241
Note. SWB = spiritual wellbeing; PTSS = perceived tangible social support; BF = benefit finding; HRQOL = health-related quality of life.
Demographics and Descriptive Statistics in the LBCS Group
27


In the Latina breast cancer survivors (LBCS) group (n = 105), participants’ ages ranged from 33 to 83 (M= 59.19; SI) = 10.64). Participants’ education level varied across the four categories, such that 25.7% had less than a high school education, 21% graduated high school, 34.3% had some college education, and 19% had at least a college degree. Most participants were married or living as married (71%). More participants were originally diagnosed with in situ or cancer (65%) compared to regional, regional NOS, or distant cancer (28%). With regard to comorbidities, about 45% of the LBCS had a Charlson Comorbidity Index (CCI) score of 0. About 25% had a CCI score of 1 or 2, and about 31% had a CCI score of 3 - 14. Table 1 (above) provides an overview of the demographic information for the LBCS group. Table 2 (above) provides an overview of the descriptive information for the LBCS group.
Demographics and Descriptive Statistics in the NLWBCS Group
In the NLWBCS group (n = 255), participants’ ages ranged from 43 to 85 (M- 62.27; SI) = 9.59 ). Participants’ education levels were such that 2.7% had less than a high school education, 16.9% graduated high school, 36.9% had some college education, and 43.5% had at least a college degree. Most participants were married or living as married (76.1%). More participants were originally diagnosed with in situ or cancer (69.8%) compared to regional, regional NOS, or distant cancer (24.7%). With regard to comorbidities, about 58% of the NLWBCS had a CCI score of 0. About 27.5% had a CCI score of 1 or 2, and about 14.5% had a CCI score of 3 - 14. Table 1 (above) provides an overview of the demographic information for the NLWBCS group. Table 2 (above) provides an overview of the descriptive information for the NLWBCS group.
Assumptions
28


Analyses involved the Whole Sample of breast cancer survivors (N = 360), which included LBCS group (n = 105) and NLWBCS group (// = 255). The evaluation of model assumptions for ANCOVA and hierarchical multiple linear regression revealed that there was no multicollinearity (i.e., correlation above .8) among any of the variables, including predictor, outcome, and covariate variables. Analysis of residuals included the calculation of Mahalanobis distance, which revealed a few outliers in the data; however, the outliers did not have a major impact on the data. The sample sizes were large enough and parameter estimation involved robust standard errors. There did not appear to be any patterns in the data with regard to multivariate outliers and missing data, and no variable exhibited more than 5 percent of missing cases. Assumptions for homoscedasticity (for the hierarchical multiple linear regression analyses) and homogeneity of variance (for the ANCOVA analyses) were not always met; however, these findings did not impact the respective analyses. Additionally, heterogeneity of the variances might reflect differences in the data based on race/ethnicity, which is a variable observed specifically in the ANCOVA analyses. Thus, in order to maintain the integrity of the data, transformations were not conducted.
Bivariate Correlation Analyses with the Whole Sample
Pearson correlation analyses were run among the covariates and primary variables in the Whole Sample. As described above, the analyses did not reveal any multicollinearity. However, the analyses did reveal several significant relationships. Race/ethnicity (1 = NLW; 2 = Latina) was negatively associated with age (p < 0.01), education ( p < .001), and health-related quality of life (HRQOL) (p < 0.01), indicating that LBCS were lower then NLWBCS on all three of these variables. Race/ethnicity was positively correlated with comorbidities (p < .01) and benefit finding (BF; p < .001). Age was positively correlated with marital status (p < .01), comorbidities
29


(p < .001), and spiritual wellbeing (SWB; p < .05). Education was negatively correlated with comorbidities (p < .001) and BF (p < .01). Education was positively correlated with HRQOL (p < .05). Marital status was negatively correlated with tumor stage at diagnosis (p < .01), perceived tangible social support (PTSS) (p < .001), and HRQOL (p < .01). Comorbidities were negatively correlated with HRQOL (p < .001). Among the primary variables, SWB was positively correlated with HRQOL (p < .001) and BF (p < .001). BF was positively correlated with HRQOL (p < .01). Results of the bivariate correlations among covariates and primary variables in the Whole Sample (N = 360) are presented in Table 3 below.
Table 3
Bivariate Correlations of Covariates and Primary Variables in the Whole Sample (N = 360)
1 2 3 4 5 6 7 8 9 10
1. Race/ _ - - .06 -.04 .17 .07 -.10 .20 - 18**
Ethnicity _ .14** .35**
2. Age -.07 .14 -.00 .21** .12 -.08 -.01 .09
** *
3. Educatio -.02 -.05 - .01 .09 - 16** .11*
n .19**
4. Marital - .04 -.06 - .00 _ 27**
Status .14** .20**
5. Tumor -.01 -.02 -.03 .03 -.00
stage at diagnosis
6. CCI -.06 -.10 .02 -.26**
score
7. SWB -.01 .33** .67**
8. PTSS .00 .10
9. BF . 14**
10. HRQOL
Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; PTSS = perceived tangible social support; BF = benefit finding; HRQOL = health-related quality of life; *p < .05; **p < .01
Bivariate Correlations with the LBCS Group
The bivariate correlations with the LBCS group revealed several significant relationships. Age was positively correlated with marital status (p < .05) and comorbidities (p < .05). Age was
30


negatively correlated with education (p < .05). Marital status was negatively associated with HRQOL (p < .01). Comorbidities were negatively correlated with HRQOL (p < .001). SWB was positively correlated with BF (p < .001) and HRQOL ( p < .001). Results of the bivariate correlations among covariates and primary variables within the LBCS group (n = 105) are presented in Table 4.
Table 4
Bivariate Correlations of Covariates and Primary Variables in the LBCS Group (n = 105)
1 2 3 4 5 6 7 8 9
1. Age _ -.23* .22* .06 .21* -.12 -.10 -.15 -.16
2. Education -.01 -.10 -.18 -.03 .12 -.11 .02
3. Marital Status -.08 .06 -.11 -.18 .02 -.26**
4. Tumor stage at diagnosis — .04 -.02 -.04 .05 .02
5. CCI score -.13 -.10 -.10 -.40**
6. SWB -.04 .34* 59**
7. PTSS -.01 .07
8. BF .16
9. HRQOL
Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; PTSS = perceived tangible social support; BF = benefit finding; HRQOL = health-related quality of life;
*p < .05; **p < .01
Group Differences on Covariates
Chi-squared analyses demonstrated significant group differences on education [x2(3) = 54.623, p < .001] and comorbidities (i.e., CCI scores) [x2(2) = 12.514,p = .01], such that NLWBCS had significantly more education and fewer comorbidities compared to LCBCS. However, there were no drastically large differences between the LBCS and NLWBCS groups on the covariates (see Tables 5 and 6 below); therefore, all 5 covariates were maintained across the main analyses (i.e., ANCOVA and hierarchical regression models).
31


Table 5
One-Way Analysis of Variance (AN OVA) of Racial/Ethnic Differences on the Continuous Covariate Age
Variable df F Effect Size (partial if)
Age 1 7.21** .02
Note. **p < .01
Table 6
Tests of Racial/Ethnic Differences on Categorical Covariates
Race/Ethnicity
Variables LBCS (n) NLWBCS (n) Df
Education 105 255 54.62*** 3
Marital Status 105 255 1.23 1
Tumor stage at diagnosis 105 255 1.08 2
CCI score 105 255 12.51** 2
Note. LBCS = Latina breast cancer survivors; NLWBCS = Non-Latina White breast cancer survivors;
*p < .05; **p < .01; ***p < .001.
Group Differences on Primary Variables (Aim 1)
ANCOVA analyses for Aim 1 tested whether there were group differences on the 4 primary variables of interest when controlling for all 5 covariates (i.e., age, education, marital status, tumor stage at diagnosis, and comorbidities). Results showed that LBCS had higher mean scores on the SWB [F (1, 356)= 4.99, p < .05] and BF [F (1, 342) = 8.77, p < .01] compared to the NLWBCS. The ANCOVA analyses showed no significant differences between the LBCS and NLWBCS mean scores on PTSS [F (1, 353) = 1.54, ns] or HRQOL [F (1, 348) = 2.96, ns] (see Table 7 on page 33).
32


Table 7
One-Way Analysis of Covariance (ANCOVA) of Racial/Ethnic Differences on Primary Variables with Sociodemographic and Medical Covariates
Variable Df F Effect Size (partial rf)
SWB 1 4.99* .01
PTSS 1 1.54 .00
HRQOL 1 2.96 .01
BF 1 8.77** .03
Note. SWB = spiritual wellbeing; PTSS = perceived tangible social support; HRQOL = health-related quality of life; BF = benefit finding; *p < .05; **p < .01.
Hierarchical Multiple Regression Analyses (Aims 2-3) with the LBCS Group
Hierarchical multiple regression analyses for Aims 2-3 examined hypothesized effects within the LBCS group on the primary variables and controlling for covariates. SWB was positively associated with HRQOL when controlling for all 5 covariates [F (6, 94) = 23.551 ,p < .001], SWB was, as hypothesized, a strong predictor of HRQOL in the LBCS [/! = .640, p < .001). Marital status and comorbidities were additionally found to be independent predictors of HRQOL [/? = -. 177, p < .05 and = - .289, p < .001, respectively] in the LBCS (see Table 8 below).
Table 8
Results of Hierarchical Multiple Regression for SWB as a Predictor of HRQOL when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105)
Variables B SE(B) ~j} If AR2 F
L0 39 23.55***
Age .03 .11 .02
Education .31 1.09 .02
Marital Status -6.46 2.47 -.18*
Tumor stage .00 .00 .03
at diagnosis CCI score -5.60 1.31 _ 29***
SWB 1.32 .14
Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; HRQOL = health-related quality of life; Degrees of freedom of F = (6, 94); *p < .05; * * */; < .001.
33


Similarly, SWB was positively associated with BF when controlling for all 5 covariates
[F (6, 92) = 2.828, p < .05], SWB was, as hypothesized, a strong predictor of BF in the LBCS group [ft = .316, p < .01]. However, none of the covariates independently predicted BF (see Table 9 below).
Table 9
Results of Hierarchical Multiple Regression for SWB as a Predictor ofBF when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105)
Variables B SE(B) ft IP AR2 F
46 40 2.83*
Age -.18 .12 -.15
Education -1.61 1.22 -.13
Marital Status 2.17 2.77 .08
Tumor stage at diagnosis .00 .00 .06
CCI score -.86 1.48 -.06
SWB .50 .16 32**
Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; BF = benefit finding; Degrees of freedom of F = (6, 92); *p < .05; **p < .01.
The analyses further showed that PTSS was positively associated with HRQOL when
controlling for all 5 covariates [F (6, 94) = 4.161, p < .01]. However, PTSS was not an
independent predictor of HRQOL in the LBCS group (ft = -.026, ns). Of the covariates, marital
status and comorbidities independently predicted HRQOL, [ft = -2.267, p < .05 and ft = -3.983, p
< .001, respectively] (see Table 10 below).
Table 10
Results of Hierarchical Multiple Regression for PTSS as a Predictor of HRQOL when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105) Variables B SE(B) ft IP AR2 F
41 TO 4.16**
Age -.07 .15 -.05
Education -.88 1.54 -.06
Marital Status -7.97 3.51 -.22*
Tumor stage .00 .00 .02
at diagnosis CCI score -7.31 1.84 - 38***
PTSS -.01 .40 -.00
34


Table 10 cont’d
Note. CCI = Charlson comorbidity index; PTSS = perceived tangible social support; HRQOL = health-related quality of life; Degrees of freedom of F = (6, 94); *p < .05; ***p < .001.
Analyses also revealed that PTSS was not associated with BF when controlling for all 5
covariates [F (6, 92) = .971, ns]. Similarly neither the primary variable nor any of the covariates
independently predicted BF in the LBCS group (see Table 11 below).
Table 11
Results of Hierarchical Multiple Regression for PTSS as a Predictor ofBF when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105)
Variables B SE(B) J) R2 AR2 F
!o6 Too ni
Age -.22 .13 -.18
Education -2.04 1.28 -.17
Marital Status 1.95 2.96 .07
Tumor stage at diagnosis .00 .00 .06
CCI score -1.47 1.55 -.10
PTSS -.01 .32 -.00
Note. CCI = Charlson comorbidity index; PTSS = perceived tangible social support; BF = benefit finding; Degrees of freedom of F = (6, 92); All F statistics were non-significant; *p < .05; **p <
.01; ***p < .001.
Analyses additionally showed that BF was positively associated with HRQOL [F =
3.509, p < .01) when controlling for all 5 covariates. However, BF was not a significant predictor of HRQOL (ft = 1.220, ns). Among the covariates, comorbidities were a significant predictor of HRQOL (fi = -3.577, p < .01) (see Table 12 below).
Table 12
Results of Hierarchical Multiple Regression for BF as a Predictor of HRQOL when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105)
Variables B SE(B) ~j} R2 AR2 F
49 m 3.51**
Age -.03 .15 -.02
Education -.66 1.55 -.04
Marital Status -6.84 3.51 -.19
Tumor stage at diagnosis .00 .00 .02
35


Table 12 cont’d
CCI score -6.64 1.86 -.35**
______________BF .15_______________43______________42__________________________________________
Note. CCI = Charlson comorbidity index; BF = benefit finding; HRQOL = health-related quality of life; Degrees of freedom of F = (6, 91); **p < .01; ***p < .001
36


Chapter IV
Discussion
Study findings demonstrated significant differences between LBCS and NLWBCS, which were consistent with the literature. Reports of lower education levels and higher comorbidities among LBCS compared to the NLWBCS echo findings in other studies (Yanez et al., 2011). However, despite lower education and higher comorbidities, the LBCS reported higher SWB and BF than their NLWBCS counterparts. These findings reflect the Latino Paradox, indicating that SWB and BF might serve LBCS in distinct ways during survivorship, especially long-term survivorship. Findings additionally showed that SWB was an independent predictor of HRQOL and BF, respectively. However, the study also demonstrated null findings that contradicted study hypotheses, such that BF was not a significant predictor of HRQOL, and PTSS was a not a significant predictor of either HRQOL or BF in the LBCS sample. Overall, the study findings might have important implications for the improvement or development of oncology programs and interventions with LBCS, as well as future research with LBCS.
With regard to SWB, researchers and clinicians should consider how LBCS might turn to spirituality as a coping strategy in ways different from NLWBCS, who have also been found to use spirituality as a coping strategy. In this way, high SWB might serve as a protective factor for LBCS and NLWBCS alike. Spirituality uniquely permeates the roots of Latino cultures and provides a rich texture to daily life (Hunter-Hernandez et al., 2015; Gallo et al., 2009); thus, it is possible that SWB serves a unique function for LBCS. Spirituality and religiosity can facilitate an individual's intrinsic relationship with God or a higher power, as well as her extrinsic relationships with family and community. Spirituality and religiosity
37


also have multidimensional relationships with other Latino cultural values, such that Latino individuals, in general, experience and express multiple values simultaneously (Campesino, 2009). Given all this, it is possible that SWB for LBCS significantly reinforces and is, in turn, reinforced by experiences that simultaneously promote other cultural values, such as personalismo (e.g., warmth, closeness, and empathy in relationships) and familismo (e.g., mutual commitment and loyalty among family members).
Another critical and relevant Latino cultural concept is colectivismo (i.e., collectivism). Collectivism may be fostered through spiritual activities, such that LBCS might find that spiritual activities, especially those involving a spiritual community, can allow them to lean on their relationships with the spiritual community and God. However, on an individual level, LBCS might experience peace and comfort through such spiritual coping activities as prayer. LBCS might additionally reflect upon the meaning of life and their purpose in life, both of which are concepts that individuals might ponder when facing their thoughts and feelings about the potential return of a terminal illness (Wildes, Miller, San Miguel de Majors & Ramirez, 2009). In this way, spirituality might serve as a channel through which LBCS can additionally reflect upon other aspects of their breast cancer experience from diagnosis to survivorship, including any losses or benefits they might have experienced.
BF or perceived benefit, as previously described, can serve as a coping strategy for breast cancer survivors. Study findings showed that SWB was a significant, independent predictor of BF in the LBCS sample. LBCS might be more likely to perceive benefits from the breast cancer experience if they consider how their relationships with God, family, friends, and community members might have been positively influenced. More specifically,
38


LBCS who experience higher SWB might also lean more into their faith and community than those who experience lower SWB, and as a result, LBCS with higher SWB might derive more benefit from the breast cancer experience.
Given the study findings, researchers and clinicians might consider tailoring psychosocial interventions with LBCS to potential, demographically-defined characteristics as lower education and higher comorbidities. More specifically, researchers and clinicians can work to match the literacy level of materials to the literacy level of the group.
Researchers and clinicians can also integrate information about and accommodate the tangible needs relevant to common comorbidities, such as providing healthy refreshments to diabetic patients or bariatric chairs for overweight patients. Additionally, researchers and clinicians can emphasize spirituality as a coping strategy in psychosocial interventions with LBCS. Oncology programs might further seek to establish connections with spiritual leaders and organizations within the hospital system or the community at large, in order to better expand their psychosocial services beyond traditional psychotherapy or counseling interventions, especially for LBCS in long-term survivorship who might be more focused on maintaining overall health and wellbeing than adjusting to survivorship.
With regard to null study findings, previous study findings in general BCS samples have found BF to be a significant predictor of HRQOL. However, the concept of BF is generally not well-understood, let alone in LBCS. The literature suggests that BF is a positive experience that leads to positive outcomes (Campbell & Woodgate, 2015; Pascoe & Edvardsson, 2013). However, much of this research involves NLW populations in the U.S., it is possible that BF is a uniquely American concept that might be expressed and understood differently for LBCS compared to NLWBCS.
39


Researchers and clinicians should carefully consider how individuals who traverse challenging circumstances perceive ‘benefit’ as positive or negative. For example, LBCS who adhere strongly to Christian or Catholic perspectives, might view an adverse experience, such as breast cancer as a punishment for having sinned. Some LBCS might perceive the breast cancer diagnosis as a divine invitation to repent and improve their lifestyle, interpersonal relationships, relationship to the community, as well as other areas of life. Individuals adhering to such beliefs might experience and view this type of benefit either positively or negatively. In this way, LBCS might derive benefit from the breast cancer experience through a distinctly spiritual or religious mechanism that is further influenced by the individual’s particular cultural context. Given all this, future research is needed to unpack the concept of BF among diverse, ethno-cultural populations, including LBCS. Qualitative methods might be especially helpful to better unpack the concept of BF among LBCS.
With regard to PTSS and HRQOL, the results of the ANCOVAs did not support the hypotheses or previous research, in which NLWBCS tend to report higher perceived social support and HRQOL than their minority counterparts, including LBCS (Garcia-Jimenez et al., 2014; Graves et al., 2012; Martinez-Ramos et al., 2013; Yanez et al., 2011). However, bivariate correlations showed that race was correlated with HRQOL, which suggests that differences might exist between the NLWBCS and LBCS groups. Thus, it is possible that the inclusion of covariates may have eliminated any statistically detectable differences. Alternatively, it may be possible that the lack of differences between the NLWBCS and LBCS on PTSS and HRQOL were due to higher homogeneity across the racial/ethnic groups in general, especially given that most of the participants were either married or living as married, regardless of race/ethnicity. It is possible that the experience of having a life partner
40


might inherently involve either the direct experience of tangible social support or the perception that it is available if needed. Hirschman & Bourjolly (2005) conducted qualitative interviews with 33 women with breast cancer, in which the majority of women reported that their partner and mother were their main sources of tangible social support. Regardless, the findings on PTSS help add to the literature, because breast cancer survivorship research, in general, has not yet focused on the influence of PTSS on HRQOL or BF. Future research should target a more diverse sample of participants with regard to marital status, in order to better observe any differences between the LBCS and NLWBCS on PTSS, and whether or not PTSS is associated with HRQOL or BF in single (i.e., never married, widowed, divorced) LBCS and NLWBCS.
With regard to HRQOL in the LBCS, results showed that marital status and comorbidities were significant predictors of HRQOL. Future research and interventions aiming to improve HRQOL in LBCS might consider whether or not marital status and comorbidities are unique predictors of HRQOL for LBCS compared to other racial/ethnic BCS. Unique predictors of HRQOL might also be useful for researchers and clinicians developing interventions that aim to improve HRQOL in LBCS. It is possible that marital status is in important predictor of HRQOL for LBCS due to the Latino cultural values of familismo (focus on family) andfamily /social support, because family/social support often includes partners (Anez et al., 2005; Gallo et al., 2009). In general, familismo and family /social support have been found to be protective, such that they often help to reduce risk and improve resilience and health outcomes in Latino populations (Hunter-Hemandez et al., 2015; Gallo et al., 2009). Therefore, marital or ‘partner’ status might play a key role in HRQOL for LBCS.
41


Overall, this study had several strengths and potential limitations. With regard to strengths, SUNSHINE was a population-based epidemiological study, such that the statewide cancer registries have the opportunity to capture data from virtually all persons diagnosed with and treated for cancer. State-wide cancer registries can also contain diverse data across this population. In this way, researchers were able to identify potential participants from a participant pool that represents the entire population of breast cancer survivors across the particular states of Colorado and Arizona, rather than, for example, specific hospital sites. Given that this study utilized such population-based data, study results are more likely to be generalizable to the LBCS population of Colorado and Arizona. Researchers were also able to observe previously collected data, such as demographic and medical data, across several years. Another key advantage of this study was that researchers were able to focus on long-term breast cancer survivorship. Participants completed the SUNSHINE surveys at least 6 years after their last treatment (Sedjo et al., 2013). Research has tended to focus on short-term survivorship, especially with regard to HRQOL. Therefore, the study results might particularly help improve understanding of the factors that contribute to HRQOL and BF among LBCS in long-term survivorship specifically.
However, study results should be interpreted caution due to potential limitations. A potential threat to internal validity is the possibility that some participants were influenced by the participant response bias of social desirability. Some Latina participants may have been hesitant to express negative emotions for fear of being stigmatized, discriminated socially, embarrassed, or labeled according to a mental illness (e.g., “depressed”) (Vega, 2010; Nadeem et al., 2007). Therefore, some Latina participants might have reported higher SWB and higher HRQOL on the respective measures. Latina participants might have also
42


attempted to respond favorably if they completed study interviews over the phone with the research assistants, or if a relative or friend helped them complete an interview(s). Participants might have wanted the relative or friend to view them in a positive light, or the relative or friend might have encouraged participants to respond more optimistically.
This study was unable to assess how mental health stigma may have influenced self-report measures within the interviews, because none of the measures involve questions on participants’ perspectives towards mental illness, such as depression or anxiety. In future studies, it would be worth taking account of mental health stigma within Latino communities. This could potentially be done by asking participants about their perspectives and beliefs related to mental illness, especially depression and anxiety. Such questions may be more feasibly posed through qualitative versus quantitative interviews, because participants could be invited to elaborate upon their perspectives. In contrast, quantitative measures would need to be improved. It is possible that responses to questions regarding mental health stigma within Latino communities may not be easily quantified, such as with the use of Likert-style responses. Mental health stigma within Latino communities might also be better conceptualized as a multidimensional construct.
Although it is hoped that the results will generalize to LBCS across the U.S., there may be limitations to external validity, particularly with regard to the structural component.
In terms of the population sample, the majority of Latinos living in Colorado and Arizona are Mexican or Mexican-American, which suggests that the results may not necessarily generalize to other Latino subgroups. It is also important to consider regional differences across the U.S. Because the LBCS sample is from the Southwest U.S. (i.e., Colorado and Arizona), they might exhibit differences in, for example, lifestyle, subculture, healthcare
43


access, access to food, ethic/racial diversity. Asa result, the LBCS sample may not generalize to LBCS living in other regions of the United States (e.g., metropolitan Chicago, metropolitan New York City, metropolitan Los Angeles). Additionally, the state-wide cancer registries might have a limited number of underserved and undocumented LBCS (Risendal et al., 2015), which could include those who lack or have limited access to quality healthcare, or have lower education. A potential lack of underserved, undocumented LBCS might explain the restriction in range on education levels among the LBCS. Latinos in general in the U.S. are significantly less likely to graduate from high school, let alone pursue higher education, compared to their NLW counterparts (Krogstad, 2016). However, the restriction in range on such demographic variables as education and marital status might be due to selection bias. Although the sample was a population-based sample and was derived from state-wide cancer registries, it is possible that the LBCS who agreed to participate tended to be those of, for example, higher education and either married or living as married.
Despite the limitations, however, the study findings not only support previous research on racial/ethnic disparities in breast cancer survivorship, but they might also contribute to improving our understanding of potential predictors of HRQOL and BF in LBCS. With improved understanding, researchers and clinicians might soon begin to fill the need for effective post-treatment interventions for LBCS throughout the Southwest United States and beyond.
44


REFERENCES
Anez, L. M., Paris, J., Manuel, Bedregal, L. E., Davidson, L., & Grilo, C. M. (2005).
Application of cultural constructs in the care of first generation Latino clients in a community mental health setting. Journal of Psychiatric Practice,11(4), 221-230. doi: 10.1097/00131746-200507000-00002
Antoni, M. H., Lehman, J. M., Kilbourn, K. M., Boyers, A. E., Culver, J. L., Alferi, S. M.. . Carver, C. S. (2001). Cognitive-behavioral stress management intervention decreases the prevalence of depression and enhances benefit finding among women under treatment for early-stage breast cancer. Health Psychology, 20(1), 20-32. doi: 10.1037/0278-6133.20.1.20
Ashing, K., Rosales, M., Lai, L., & Hurria, A. (2014). Occurrence of comorbidities among African-American and Latina breast cancer survivors. Journal of Cancer Survivorship, 8(2), 312-318. doi:10.1007/sll764-014-0342-x
Ashing-Giwa, K. T., Kim, J., & Tejero, J. S. (2008). Measuring quality of life among
cervical cancer survivors: Preliminary assessment of instrumentation validity in a cross-cultural study. Quality of Life Research, 17( 1), 147-157. doi: 10.1007/slll36-007-9276-3
Ashing-Giwa, K., J. Tejero, et al. (2007). Examining predictive Models of HRQOL in a
population based, multi-ethnic sample of women with breast carcinoma. Quality of Life Research, 16, 413- 428.
Askim-Lovseth, M.K. & Aldana, A. (2010). Looking beyond “affordable” health care: Cultural understanding and sensitivity—necessities in addressing the health care disparities of the U.S. Hispanic population. Health Marketing Quarterly, 27:4, 354-387, DOI: 10.1080/07359683.2010.519990
Brady, M. J., Peterman, A. H., Pitchett, G., Mo, M., & Celia, D. (1999). A case for including spirituality in quality of life measurement in oncology. Psycho-oncology, 8(5), 417-428. doi: 10.1002/(SICI) 1099-1611( 199909/10)8:5<417::AID-PON398>3.0.CO;2-4
Brady, M. J., Celia, D. P., Mo, P., Bonomi, A. E., Tulsky, D. S., Lloyd, S. R.. . Shiomoto, G. (1997). Reliability and validity of the functional assessment of cancer therapy-breast quality-of-life instrument. Journal of Clinical Oncology, 15(3), 974-986.
Braveman, P. A., Cubbin, C., Egerter, S., Williams, D. R., & Pamuk, E. (2010).
Socioeconomic disparities in health in the United States: What the patterns tell us. American Journal of Public Health, 100( SI), S186-S196. doi: 10.2105/AJPH.2009.166082
45


Bredle, J. M., Salsman, J. M., Debb, S. M., Arnold, B. J. & Celia, D. (2011). Spiritual
wellbeing as a component of health-related quality of life: the functional assessment of chronic illness therapy—spiritual wellbeing scale (FACIT-Sp). Religions, 2(11), 77-94. doi: 10.3390/rel2010077
Bronfenbrenner, U. (1994). Ecological models of human development. In International Encyclopedia of Education, Vol. 2, 2nd edition. Oxford: Elsevier.
Brookings J. B., Bolton B. (1988). Confirmatory factor analysis of the Interpersonal Support Evaluation List. American Journal of Community Psychology, 16, 137-147
Campbell-Enns, H. & Woodgate, R. (2015). The psychosocial experiences of women with breast cancer across the lifespan: a systematic review protocol. The Joanna Briggs Institute Database of Systematic Reviews Implementation Reports. 13(1): 112-21. doi: 10.11124/jbisrir-2015-1795.
Campesino, M. (2009). Spirituality Among Latinas/os Implications of Culture in
Conceptualization and Measurement. Advances in Nursing Science. 29(1): 69-81.
Carver, C. S. (2013). The Benefit Finding Scale for breast cancer. Measurement Instrument Database for the Social Science. Retrieved from www.midss.ie
Celia, D. F., Tulsky, D. S., Gray, G., Sarafian, B., Linn, E., Bonomi, A... & Brannon, J. (1993). The functional assessment of cancer therapy scale: Development and validation of the general measure. Journal of Clinical Oncology ,11(3), 570.
Celia, D., Hernandez, L., Bonomi, A. E., Corona, M., Vaquero, M., Shiomoto, G., & Baez, L. (1998). Spanish language translation and initial validation of the functional assessment of cancer therapy quality-of-life instrument. Medical Care, 36(9), 1407-1418. doi: 10.1097/00005650-199809000-00012
Chopra, I. & Kamal, K.M. (2012). A systematic review of quality of life instruments in longterm breast cancer survivors. Health and Quality ofEife Outcomes, 10( 14). doi: 10.1186/1477-7525-10-14
Cohen, S., & Hoberman, H. (1983). Positive events and social supports as buffers of life change stress. Journal of Applied Social Psychology, 13, 99-125.
Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. M. (1985). Measuring the
functional components of social support. In I. G. Sarason & B. R. Sarason (Eds.), Social support: Theory, research, and applications (pp. 73-94). The Hague, The Netherlands: Martinus Nijhoff.
46


Crookes, D. M., Shelton, R. C., Tehranifar, P., Aycinena, C., Gaffney, A. O., Koch, P.. . Greenlee, H. (2016). Social networks and social support for healthy eating among Latina breast cancer survivors: Implications for social and behavioral interventions. Journal of Cancer Survivorship, 10(2), 291-301. doi: 10.1007/sl 1764-015-0475-6
Davis, L. E., Engel, R. J., & SpringerLink (Online service). (2011). “Chapter 2: Hispanics.” Measuring race and ethnicity. New York: Springer, doi: 10.1007/978-1-4419-6697-1_2
Gallo, L. C., Penedo, F. J., Espinosa de los Monteros, Karla, & Arguelles, W. (2009).
Resiliency in the face of disadvantage: Do Hispanic cultural characteristics protect health outcomes? Journal of Personality, 77(6), 1707-1746. doi: 10. Ill 1/j. 1467-6494.2009.00598.x
Garcia-Jimenez, M., Santoyo-Olsson, J., Ortiz, C., Lahiff, M., Sokal-Gutierrez, K., &
Napoles, A. M. (2014). Acculturation, inner peace, cancer self-efficacy, and self-rated health among Latina breast cancer survivors. Journal of Health Care for the Poor and Underserved, 25(4), 1586-1602. doi:10.1353/hpu.2014.0158
Garcfa-Jimenez, M. (2013). Culture and Self-rated Health Among Latina Breast Cancer Survivors: The Role of Cancer Self-efficacy and Spiritual Wellbeing (Master’s thesis). Retrieved from eScholarship University of California
Graves, K. D., Jensen, R. E., Canar, J., Perret-Gentil, M., Leventhal, K., Gonzalez, F... Mandelblatt, J. (2012). Through the lens of culture: Quality of life among Latina breast cancer survivors. Breast Cancer Research and Treatment, 136(2), 603-613. doi: 10.1007/s 10549-012-2291-2
Hegelson, V. S., Reynolds, K. A., & Tomich, P. L. (2006). A meta-analytic review of benefit finding and growth. Journal of Consulting and Clinical Psychology, 74(5), 797-816. doi: 10.1037/0022-006X.74.5.797
Hirschman, K. B. & Bourjolly, J. N. (2005). How Do Tangible Supports Impact the Breast Cancer Experience? Social Work in Health Care, 41:1, 17-32, DOI:
10.1300/J010v41n01_02
Hoddinott, S. N., & Bass, M. J. (1986). The Dillman total design survey method. Canadian Family Physician Medecin De Famille Canadien, 32, 2366-2368.
Hodge, D. R., Wolosin, R. J., & Bonifas, R. P. (2013). Addressing older Latinos' spiritual needs in hospital settings: Identifying predictors of satisfaction. Advances in Social Work, 14(1), 65-81.
Hunter-Hernandez, M., Costas-Munfz, R., & Gany, F. (2015). Missed opportunity:
Spirituality as a bridge to resilience in Latinos with cancer. Journal of Religion and Health, 54(6), 2367-2375. doi: 10.1007/sl0943-015-0020-y
47


IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp.
Janz, N.K., Friese, C.R., Li, Y., Graff, J.J., Hamilton, A.S., & Hawley, S.T. (2014).
Emotional wellbeing years post-treatment for breast cancer: prospective, multi-ethnic, and population-based analysis. Journal of Cancer Survivorship, 8, 131-142. doi:
10.1007/s 11764-013-0309-3
Jansen, L., Hoffmeister, M., Chang-Claude, J., Brenner, H. & Arndt, V. (2011). Benefit finding and post-traumatic growth in long-term colorectal cancer survivors: prevalence, determinants, and associations with quality of life. British Journal of Cancer, 105, 1158-1165. doi: 10.1038/bjc.2011.335
Kagawa-Singer, M., Padilla, G. V., & Ashing-Giwa, K. (2010). Health-related quality of life and culture. Seminars in Oncology Nursing,26(1), 59-67. doi: 10.1016/j.soncn.2009.11.008'
Kroenke, C. H., Quesenberry, C., Kwan, M. L., Sweeney, C., Castillo, A., & Caan, B. J.
(2013a). Social networks, social support, and burden in relationships, and mortality after breast cancer diagnosis in the life after breast cancer epidemiology (LACE) study. Breast Cancer Research and Treatment, 137(1), 261-271. doi: 10.1007/s 10549-012-2253-8
Kroenke, C. H., Kwan, M. L., Neugut, A. I., Ergas, I. J., Wright, J. D., Caan, B. J.. . Kushi,
L. H. (2013b). Social networks, social support mechanisms, and quality of life after breast cancer diagnosis. Breast Cancer Research and Treatment, 139(2), 515-527. doi: 10.1007/sift549-013-2477-2
Kroenke C. H., Michael Y., Tindle H., Gage E., Chlebowski R., Garcia L., Messina C., Manson J. E., Caan B. J. (2012) Social networks, social support and burden in relationships, and mortality after breast cancer diagnosis. Breast Cancer Res Treat 133(1): 375-385
Krogstad, J.M. (2016). 5 Facts About Latinos and Education. The Pew. Retrieved from: http://www.pewresearch.org/fact-tank/2016/07/28/5-facts-about-latinos-and-education/
Levine, E. G., Yoo, G., Aviv, C., Ewing, C., & Au, A. (2007). Ethnicity and spirituality in breast cancer survivors. Journal of Cancer Survivorship, 1(3), 212-225. doi: 10.1007/s 11764-007-0024-z
Lopez-Class, M., Perret-Gentil, M., Kreling, B., Caicedo, L., Mandelblatt, J., & Graves, K. D. (2011). Quality of life among immigrant Latina breast cancer survivors: Realities of culture and enhancing cancer care. Journal of Cancer Education, 26(4), 724-733. doi: 10.1007/s 13187-011-0249-4
48


Maly, R. C., Stein, J. A., Umezawa, Y., Leake, B., & Anglin, M. D. (2008). Racial/Ethnic differences in breast cancer outcomes among older patients: Effects of physician communication and patient empowerment. Health Psychology, 27(6), 728-736. doi: 10.1037/0278-6133.27.6.728
Martinez-Ramos, G. P., Biggs, M. J. G., & Lozano, Y. (2013). Quality of life of Latina breast cancer survivors: From silence to empowerment. Advances in Social Work, 14(1), 82-101.
Mickley, J., & Soeken, K. (1993). Religiousness and hope in Hispanic- and Anglo-American women with breast cancer. Oncology Nursing Forum, 20(8), 1171.
Nadeem, E., Lange, J. M., Edge, D., Fongwa, M., Belin, T., & Miranda, J. (2007). Does stigma keep poor young immigrant and U.S.-born black and Latina women from seeking mental health care? Psychiatric Services (Washington, D.C.), 58(12), 1547-1554. Napoles, A. M., Ortiz, C., O'Brien, H., Sereno, A. B., & Kaplan, C. P. (2011). Coping resources and self-rated health among Latina breast cancer survivors. Oncology Nursing Forum, 38(5), 523. doi: 10.1188/1 l.ONF.523-531
Ortega, A. N., Rodriguez, H. P., & Vargas Bustamante, A. (2015). Policy dilemmas in Latino health care and implementation of the affordable care act. Annual Review of Public Health, 36(1), 525-544. doi:10.1146/annurev-publhealth-031914-122421
Pew Research Center (2014). The shifting religious identify of Latinos in the United States:
Nearly one in four Latinos are former Catholics. Pew Research Center. Retrieve from
http://www.pewforum.org/2014/05/07/the-shifting-religious-identity-of-latinos-in-
the-united-states/
Pinheiro, P. S., Williams, M., Miller, E. A., Easterday, S., Moonie, S., & Trapido, E. J.
(2011). Cancer survival among Latinos and the Hispanic Paradox. Cancer Causes & Control, 22(4), 553-561. doi:10.1007/s 10552-011-9727-6
Pascoe, L., Edvardsson, D., Umea universitet, Institutionen for omvardnad, & Medicinska fakulteten. (2013). Benefit finding in cancer: A review of influencing factors and health outcomes. European Journal of Oncology Nursing: The Official Journal of European Oncology Nursing Society, 17(6), 760. doi: 10.1016/j.ejon.2013.03.005
Patel, M. I., Schupp, C. W., Gomez, S. L., Chang, E. T. & Wakelee, H. A. (2013). How do social factors explain outcomes in non-small-cell lung cancer among Hispanics in California? Explaining the Hispanic Paradox. Journal of Clinical Oncology. 31(28). doi: 10.1200/JCO.2012.48.6217
Rasheed, M. N., Rasheed, J. & Marley, J. A. (2011). Cultural values in relation to family structure. Family Therapy: Models and Techniques, (pp. 111-115). Thousand Oaks, CA: Sage Publications, Inc.
49


Risendal, B. C., Sedjo, R. L., Giuliano, A. R., Vadaparampil, S., Jacobsen, P. B., Kilbourn,
K.. . Byers, T. (2015). Surveillance and beliefs about follow-up care among long-term breast cancer survivors: A comparison of primary care and oncology providers. Journal of Cancer Survivorship, 10(1), 96-102. doi: 10.1007/sl 1764-015-0454-y
Ruiz, J. M., Campos, B. & Garcia, J. J. (2016). Special issue on Latino physical health:
Disparities, paradoxes, and future directions. (2016). Journal ofLatina/o Psychology, 4(2), 61-66. doi: 10.1037/lat0000065
Schwartz, S. J., Unger, J. B., Zamboanga, B. L., & Szapocznik, J. (2010). Rethinking the concept of acculturation: Implications for theory and research. American Psychologist, 65(4), 237-251. doi: 10.1037/a0019330
Sedjo RL, Hines LM, Byers T, Giuliano AR, Marcus A, Vadaparampil S, Jacobsen P,
Kilbourn K, Feser WJ, Risendal BC (2013). Long-term weight gain among Hispanic and non-Hispanic White women with and without breast cancer. Nutrition and Cancer. 65: 34-42. PMID 23368911 DOI: 10.1080/01635581.2013.741750
Siegel, R. L., Fedewa, S. A., Miller, K. D., Goding-Sauer, A., Pinheiro, P. S., Martinez-
Tyson, D., & Jemal, A. (2015). Cancer statistics for Hispanics/Latinos, 2015. CA: A Cancer Journal for Clinicians, 65(6), 457-480. doi: 10.3322/caac.21314
Tammemagi, C.M., Nerenz, D., Neslund-Dudas, C., Feldkamp, C. & Nathanson, D. (2005). Comorbidity and survival disparities among black and white patients with breast cancer. The Journal of the American Medical Association. 294(14).
Tomich, P. L., & Helgeson, V. S. (2004). Is finding something good in the bad always good? benefit finding among women with breast cancer. Health Psychology, 23(1), 16-23. doi: 10.1037/0278-6133.23.1.16
Uchino, B. N. (2004). Social support and physical health: Understanding the health consequences of relationships. New Haven: Yale University Press.
University of Manitoba (2016). Concept: Charlson Comorbidity Index. Retrieved from: http://mchp-appserv.cpe.umanitoba.ca/viewConcept.php7conceptIDM098
U.S. Census Bureau (2015). Hispanic Heritage Month 2015. Profile American Facts for Features. Retrieved from:
https://www.census.gov/content/dam/Census/newsroom/facts-for-features/2015/cbl5-ff 18 hispanic heritage month.pdf
Vega, W. (2010). Addressing stigma of depression in Latino primary care patients. General Hospital Psychiatry, 32(2), 182-191. doi:10.1016/j.genhosppsych.2009.10.008
50


Viruell-Fuentes, E. A., & Schulz, A. J. (2009). Toward a dynamic conceptualization of social ties and context: Implications for understanding immigrant and Latino health. American Journal of Public Health, 99(12), 2167-2175. doi: 10.2105/AJPH.2008.158956
Weaver, K. E., Llabre, M. M., Lechner, S. C., Penedo, F., & Antoni, M. H. (2008).
Comparing unidimensional and multidimensional models of benefit finding in breast and prostate cancer. Quality of Life Research, 17(5), 771-781. doi: 10.1007/sl 1136-008-9348-z
Wheeler, S.B., Reeder-Hayes, K.E. & Carey, L.A. (2013). Disparities in breast cancer treatment and outcomes: Biological, social, and health system determinants and opportunities for research. The Oncologist, 18, 986-993. doi: http://dx.doi.org/10.1634/theoncologist.2013-0243
Wildes, K. A., Miller, A. R., San Miguel de Majors, S. & Ramirez, A. G. (2009). The religiosity/spirituality of Latina breast cancer survivors and influence on health-related quality of life. Psychooncology. 18(8): 831-840. doi: 10.1002/pon. 1475
Wu, A. H., Kurian, A. W., Kwan, M. L., John, E. M., Lu, Y., Keegan, T. H. M.. . Vigen, C. (2015). Diabetes and other comorbidities in breast cancer survival by race/ethnicity: The California breast cancer survivorship consortium (CBCSC). Cancer Epidemiology, Biomarkers & Prevention: A Publication of the American Association for Cancer Research, Cosponsored by the American Society of Preventive Oncology, 24(2), 361-368. doi: 10.1158/1055-9965.EPI-14-1140
Yanez, B., Thompson, E. H., & Stanton, A. L. (2011). Quality of life among Latina breast cancer patients: A systematic review of the literature. Journal of Cancer Survivorship, 5(2), 191-207. doi: 10.1007/sl 1764-011-0171-0
51


Full Text

PAGE 1

PREDICTING HEALTH RELATED QUALITY OF LIFE AND BENEFIT FINDING IN LATINA BREAST CANCER SURVIVORS By LAUREN LANGUIDO Post Baccalaureate Certificate , Northwestern University, 2013 B.A., DePaul University, 2009 A t hesis s ubmitted to the Faculty of the Gr aduate School of the University of Colorado Denver in partial fulfillment of the r equirements for the degree of Master of Arts Clinical Health Psychology Program 2017

PAGE 2

! ! ! ii © 2017 LAUREN LANGUIDO ALL RIGHTS RESERVED

PAGE 3

! ! ! iii This thesis for the M aster of Arts degree by Lauren Languido has been approved for the Clinical Health Psychology Program by Kristin Kilbourn , Chair Betsy Risendal Evelinn Borrayo Krista Ranby Date: December 16, 2017

PAGE 4

! ! ! iv Lauren Languido (B.A. ) Predicting Health Related Qu ality of Life and Benefit Finding in Latina Breast Cancer Survivors Thesis directed by: Assistant Professor Kristin Kilbourn ABSTRACT As the Latino population in the United States (U.S.) continues to grow, so do health disparities, especially for those fa cing such serious illness as cancer. About 91.9 of every 100,000 Latinas were diagnosed with breast cancer between 2008 and 2012. Early detection and treatments have improved and have led to increased survivorship among Latinas . Survivorship involves uniqu e challenges that may compound the challenges that many Latinas already face on a daily basis (e.g., socioeconomic challenges, limited or no health insurance, limited access to healthcare ). Research has begun to evaluate how cultural and psychosocial facto rs influence health outcomes among Latina breast cancer survivors (LBCS). Data for the proposed study was derived from a sample of LBCS ( n = 105) who participated in a multi site, longitudinal, follow up study, entitled "Survivorship Update Network to Sout hwest Hormone, Insulin, Nutrition, and Exercise [SUNSHINE]," which took place in Arizona and Colorado between April 2007 and July 2008. The study aimed to extend previous literature and examine potential predictors of health related quality of life (HRQOL) and benefit finding (BF) in LBCS . S tudy findings showed that spiritual wellbeing independently predicted HRQOL and BF , and BF independently predicted HRQOL over and above age, marital status, education level, tumor stage and diagnosis, and comorbidities. However, p erceived tangible social support did not predict HRQOL or BF in our sample. It is hoped that study findings will h elp improve understanding of how cultural factors might influe nce health outcomes among LBCS , as well as how intervention programs c an be tailored to this demographic ally defined group .

PAGE 5

! ! ! v The form and content of this abstract are approved. I recommend its publication. Approved: Kristin Kilbourn

PAGE 6

! ! ! vi TABLE OF CONTENTS CHAPTER I. INTRODUCTION..... ................................................. ........................ .......... ...... ..... . ...... 1 13 Overview .. ........................................................ ........................... ........................ ............ .... 1 Breast cancer incidence and prevalence among Latinas ÉÉÉÉ ... .............................. 1 2 Health disparities and LBCS....... ..................... .................. ................... .. ......... ................ 2 3 Theoretical Context ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ... ..... ........... .......... ............ 3 6 Health dispar ities and LBCS.................................................................... ................. ....... 6 8 P erceived social support Among LBCS ÉÉÉÉÉÉÉÉÉ............... .......... ............. 8 9 Benefit Finding and Cultural Factors Among LBCS ÉÉÉÉ... ............ . ... ................. 9 11 The Present Study .................................................................. ........................... ........... 11 13 II. METHODS ..................................................... ........................... . ... ........... .................... 14 22 Sample ........................................................................................ . ...................... ................ 14 Procedure .................................................................. .................. ............... ................ ... .... 15 Measures ...... ..... . .............................. .......................... ................ ........................... . ....... 15 21 Statistical Analysis ........................................ .................... ........................... ..... .......... 21 22 III. RESULTSÉÉÉ .................. .................... .............................................. ........... .......... 23 33 Demographics and Descriptive Statistics in the Whole Sample ..... .................. ............ 23 25 Demographics and Descriptive Statistics in the LBCS Group........................ ... .......... 25 26 Demographics and Descriptive Statistics in the NLWBCS Group.................. . ........... . . ... 26 Assumptions ........ ................................................................................................... ..... 26 27 Bivariate Correlation Analyses with the Whole Sample ........................ .... ........ .... ...... 27 28 Bivariate Correlation Analyses with the LBCS Group .............. .................... ........ ...... 28 29

PAGE 7

! ! ! vii Group Differences on Covariates .................................................................... ....... ...... 29 30 Group Differences on Primary Variables ( Aim 1) ........................ ...... ................... ....... .... 30 Hierarchical Multiple Regression Analyses ( Aims 2 3) with the LBCS Group . ... ...... 30 33 IV. DISCUSSIONÉÉÉÉ .......................................................... .................... . ........... . .... 34 41 V. REFERENCES .. ............................................... .................... .......................... ........ ...... 42 51

PAGE 8

! ! ! viii LIST OF TABLES TABLE 1 . Demographic Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), & NLWBCS Group (n = 255) ÉÉÉÉ É É É É É É É É É É É É É É É É É É É É É É 25 7 2. Des criptive Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), & NLWBCS Group (n = 255) ) ÉÉÉÉ É É É É É É É É É É É É É É É É É É É É É É É É É É . .. 27 3 . Bivariate Correlations of Covariates and Primary Variables in the Whole Sample (N = 360) . . ...................................................... .......................................................................... .............. 3 0 4 . Bivariate Correlations of Covariates and Primary Variables i n the LBCS Group (n = 105) É 3 1 5 . One Way Analysis of Variance (ANOVA) of Racial/Ethnic Differences on the Continuous Covariate Age ) ÉÉÉÉ É É É É É É É É É É É É É É É É É É É É É É . .. É É É É É 32 6 . Tests of Racial/Ethnic Differences on Categorical Covariates É É É É É É É É É É .. É É 32 7 . One Way Analysis of Covariance (ANCOVA) of Racial/Ethnic Differences on Primary Variables when controlling for Sociodemographic and Medical Covariates É É É É É É É É 3 3 8 . Results of Hierarchical Multiple Regression for SWB as a P redictor of HRQOL when controlling for Sociodemographic and Medical Covariates in LBCS (n = 105) É É É É É É .. . 3 3 9 . Results of Hierarchical Multiple Regression for SWB as a Predictor of BF when controlling for Sociodemographic and Medical Covariates in LBCS (n = 105) É É É É É É É É É É É 3 4 1 0 . Results of Hierarchical Multiple Regression for PTSS as a Predictor of HRQOL when controlling for Sociodemographic and Medical Covariates in LBCS (n = 105) É É É É É É 34 5 1 1 . Results of Hierarchical Multiple Regression for PTSS as a Predictor of BF when controlling for Sociodemographic and Medical Covariates in LBCS (n = 105) É É É É É É É É É É É .. 3 5 1 2 . Results of Hierarchical Multiple Re gression for BF as a Predictor of HRQOL when controlling for Sociodemographic and Medical Covariates in LBCS (n = 105) É É É É É É 35 6

PAGE 9

! ! ! 1 CHAPTER I INTRODUCTION Overview Latinos make up the largest and fastest growing ethnic minority population in the United States (U.S. ; Braveman et al., 2010; Graves et al., 2012). There are currently over 55 million Latinos living in the U.S., a number that is projected to reach 119 million by 2060 (U.S. Census Bureau, 2015). As the Latino population continues to grow, significant socio economic challenges and health issues that affect Latinos also increase (Ortega, Rodriguez, & Vargas Bustamante, 2015). In general, Latinos experience higher rates of poverty and adverse health outcomes, and often lack access to health care (Askim Lovseth & Aldana, 2010; Gallo et al., 2009; Graves et al., 2012; Vega, Rodriguez, & Gruskin, 2009). Latinos often live without proper treatment for health issues until their illness es have progressed. This is often the case for Latinos with cancer, especially Lati na breast cancer patients. Latinas are often diagnosed at later stages of breast cancer and late stage treatment typically involves more aggressive and costly i nterventions and poorer prognose s (Askim Lovseth & Aldana, 2010; Ortega et al., 2015). Latina wo men with breast cancer face particularly significant health disparities compared to their Non Latino White ( NLW ) counterparts (Yanez et al., 2011). Breast cancer incidence and prevalence among Latinas Although breast cancer incidence rates are lower among Latina women, breast cancer is the most common cancer and the most common cause of cancer related deaths among Latinas. (American Cancer Society, 2015; Center for Disease Control and Prevention, 2015; Siegel et al., 2015). Between 2008 and 2012, approxima tely 91.9 of every 100,000 Latinas were diagnosed with breast cancer, and about 14.5 of every 100 Latinas died of br east cancer (Siegel et al.,

PAGE 10

! ! ! 2 2015 ). More recently, Latinas have had greater exposure to better screening practices and access to improved tre atment programs. Improved care has likely led to higher breast cancer survival rates and longevity among Latina breast cancer survivors (LBCS) both in general and in comparison to Non Latina White breast cancer survivors ( NLWBCS; Graves et al., 2012). Howe ver, LBCS in general continue to face significant challenges to their overall health, including lower education and socioeconomic status, lack of health insurance, limited access to healthcare, and poorer quality of life ( QOL ) compared to NLWBCS (Ruiz, Cam pos & Garcia, 2016; Pinheiro et al., 2011; Yanez et al., 2011). Health disparities and LBCS As mentioned above, research has shown that LBCS experience poorer overall health, quality of life (QOL), and health related quality of life (HRQOL) compared to NL WBCS (Garcia Jimenez et al., 2014; Graves et al., 2012; Martinez Ramos et al., 2013; Yanez et al., 2011). In general, QOL is a multidimensional concept that reflects an individual's sense of wellbeing derived from his/her satisfaction or dissatisfaction ac ross a variety of life domains. Such life domains include physical, functional, psychological, spiritual, and/or social wellbeing. QOL is a broad term that may involve distinct cultural perspectives on wellbeing and may account for unique cultural contexts and value systems that may vary across cultural and ethnic groups (Kagawa Singer, Padilla & Ashing Giwa, 2010; Sammarco & Konecny, 2008). HRQOL is similar to QOL in that HRQOL involves the same range of wellbeing domains and may be defined and expressed d ifferently across cultural and ethnic groups. However, HRQOL differs from general QOL in that it elucidates the individual's perception of how his/her health and disease status impacts his/her overall wellbeing (Kagawa Singer et al., 2010).

PAGE 11

! ! ! 3 Ashing Giwa et al. (2007) and Maly et al. (2008) conducted studies that demonstrated that LBCS experience poorer HRQOL compared to their African American and NLW counterparts. In a systematic review, Yanez et al. (2011) evaluated 22 studies comparing QOL among LBCS com pared with other ethnic/racial groups. LBCS, in general across the studies, tended to report significantly poorer QOL compared to NLWBCS and experienced the largest differences across physical health, social wellbeing, and sexual health domains. In additio n, a qualitative study revealed that LBCS primarily struggle with psychological concerns (e.g., crying and feeling sad, anxious, and irritable), followed by social functioning and spiritual and existential concerns. In another qualitative study examining QOL among LBCS alone and in comparison to other ethnic groups, LBCS participants discussed the harmful side effects of treatment whereas NLWBCS discussed positive aspects of the breast cancer experience. Some quantitative studies confirmed findings that LB CS experienced poorer psychological and physical health compared to their NLWBCS counterparts. LBCS additionally reported more disruptions to social life and relationships than NLWBCS. Differences between LBCS and NLWBCS on outcome measures of QOL as well as psychological and physical health persisted even when researchers accounted for differences in treatment type, socioeconomic status (SES), and other sociodemographic variables. However, in the same vein, other studies found that LBCS experienced higher QOL and, as mentioned above, lower mortality rates and increased longevity compared to NLWBCS (Yanez et al., 2011). Theoretical Framework HRQOL and Contextual/Cultural Factors

PAGE 12

! ! ! 4 Differences between LBCS and NLWBCS on health outcomes such as QOL may be due t o differences in situational, contextual , and cultural factors (Garcia Jimenez, 2013; Kawaga Singer et al., 2010). Ashing Giwa's model of HRQOL highlights culture as a macro component of HRQOL. Ashing Giwa's model of HRQOL might be better understood throug h the lens of Br onfenbrenner's ecological model, especially because Ashing Giwa employs a complex, contextual framework to explain HRQOL. This contextual framework involves the interrelationships among macro level components (e.g., cultural, socio ecologic al, demographic, health care systems factors, etc.) and micro or individual level components (e.g., health status, disease characteristics, health efficacy, level of functioning, etc.) (Bronfenbrenner, 1994; Kagawa Singer et al., 2010; Ashing Giwa et al., 2007) . The model's contextual framework also involves the unique and interacting influences of macro and micro level components on the various dimensions of wellbeing . Ashing Giwa's model of HRQOL included the same dimensions of HRQOL as other measures o f HRQOL and general QOL. Ashing Giwa's model of HRQOL focused on the following dimensions of HRQOL: physical wellbeing , functional wellbeing , psychological and emotional wellbeing , social wellbeing , spiritual wellbeing, and sexual wellbeing (Kagawa Singer et al., 2010; Ashing Giwa et al., 2007). LBCS in the U.S. generally face high adversity such as lower income, education, and employment opportunities, as well as higher rates of discrimination, and limited or lack of access to health insurance and quality care (Ruiz et al., 2016). However, despite high adversity, LBCS tend to experience lower mortality rates and increased longevity compared to NLWBCS. This comparative advantage is an important example of the Latino Health Paradox (Pinheiro et al., 2011; Vi ruell Fuentes & Schulz, 2009). The Latino Health Paradox is an epidemiological phenomenon in which Latinos , in general , overcome adversity and experience more health

PAGE 13

! ! ! 5 advantages compared to Non Latino Whites with regard to physical and psychological health outcomes. Recently, research has shown that Latinos' comparative advantages may be due to cultural characteristics such as strong social integration and cohesion, and a strong emphasis on spirituality and/or religiosity (Gallo et al., 2009; Patel et al., 2 013; Ruiz et al., 2016). The Latino Health Paradox and Cultural Values Some research suggests that adherence to some Latino cultural values might lead to better health outcomes (Gallo et al., 2009; Patel et al., 2013; Ruiz et al., 2016). For example, more ethnically dense Latino neighborhoods are associated with better physical and psychological health for both Latinos and Non Latinos (Ruiz et al., 2016). Latino cultural values related to interpersonal relationships are commonly cited in the literature. Su ch values include familismo (focus on family), personalismo (personal versus institutional relationships), respeto (mutual and reciprocal respect), confianza (trust and intimacy in relationships), collectivism , and family/social support (A–ez et al., 2005; Gallo et al., 2009). Similarly, religiosity and/or spirituality are Latino cultural values that are often viewed and experienced as enmeshed concepts. Religiosity and/or spirituality are often understood as relating to the interpersonal relationships that one creates with a religious and/or spiritual community as well as the intrapersonal experience of relating to a higher power (Hunter Hern‡ndez et al., 2015). Religiosity and spirituality as well as social support have been especially highlighted in resea rch with LBCS as protective cultural factors. Such factors have also been found to reduce risk and improve both resilience and health outcomes among Latinos in general (Hunter Hern‡ndez et al., 2015; Gallo et al., 2009). However, Latinos might experience reduced resilience and higher risk of adverse health outcomes when they lose or lack protective Latino cultural factors. One such Latino cultural factor is social support . LBCS who do not have access to social support may experience a higher

PAGE 14

! ! ! 6 risk of recurr ence and comorbid physical and psychological health issues in the survivorship period (Schwartz et al., 2013; Rasheed et al., 2011; N‡poles et al., 2011). A history of cancer has been found to be associated with breast cancer specific mortality significant ly more often in Latinas than any other ethnic/racial group (Wu et al., 2014). In general, comorbidities such as diabetes and obesity have been found to be associated with breast cancer specific mortality regardless of ethnic/racial group (Connor et al., 2 016; Wu et al., 2014). Connor et al (2016) found that LBCS were more likely to have a history of diabetes and to be classified as obese compared to their NLW counterparts, which may partially explain why comorbid diabetes and obesity are more highly associ ated with breast cancer specific mortality in LBCS. In addition, Ashing et al. (2014) assessed comorbidities in breast cancer survivorship among LBCS ( n = 232) and African American BCS ( n = 88) given that both groups tend to face significant health dispar ities. The researchers found that LBCS reported twice as many headaches and migraines, as well as significantly more osteoporosis than African American BCS (Ashing et al., 2014). Given the foregoing weight of the evidence in the literature , the factors tha t contribute to health outcomes and overall QOL in LBCS are not well understood. As a result, Latina breast cancer survivorship and the mechanisms contributing to risk and resilience are of prime interest to clinical researchers (Martinez Ramos, Biggs & Lo zano, 2013). More specifically, research is needed to further explore the relationships between specific cultural values and health outcomes (Ruiz et al., 2016). Spiritual Wellbeing among LBCS Spiritual wellbeing (SWB) is conceptualized as a multidimens ional construct that may encompass both religious and non religious derivations of faith and inner peace (Garc’a Jimenez et al., 2014). Spirituality generally refers to an individual's internal experience of faith, hope, and

PAGE 15

! ! ! 7 meaning, whereas religiosity re fers to the behaviors and rituals in which an individual engages and that are based on organized religion. However, as mentioned above, many individuals may experience spirituality and religiosity as enmeshed concepts (Levine, Yoo, Aviv, Ewing & Au, 2007). For many Latinos, spirituality is considered intrinsic to Latino culture (Yanez et al., 2011; Gallo et al., 2009). Latinos in general may adhere to a broad range of spiritual and religious beliefs, values, and practices that may vary according to gender, geographic region, specific Latino sub culture (e.g., Mexican, Guatemalan, Cuban, Ecuadorian, Spanish, etc.) (Hodge et al., 2013). Regardless of such differences, Latinos as a group in the U.S. are recognized for their high rates of spirituality and religi osity compared to NLW. For example, about 55 percent of the 35.4 million Latino adults in the U.S. consider themselves Catholic, followed by other Christian religions, and then by non Christian religions (Pew Research Center, 2008). Unlike NLW in the U.S. who tend to experience spirituality and religion as separate from daily life, Latino individuals and communities experience spirituality and religion as infused in daily life. In this way, Latino culture embraces and influences spiritual and religious int erpretation and expression among Latinos (Hodge at al., 2013). Latinos are known to engage in spiritual and religious coping when dealing with a variety of difficult life situations including cancer diagnosis, treatment, and survivorship (Yanez et al., 201 1). Gallo et al. (2009) highlighted several studies that demonstrated that higher levels of spiritual and religious coping were associated with higher levels of self rated satisfaction with life, a reduction of harmful health behaviors, and higher treatmen t seeking among substance users of diverse populations including Latinos (Gallo et al., 2009). With regard to LBCS, Levine et al. (2007) used a qualitative approach including open ended questions to evaluate coping styles among diverse BCS. Based

PAGE 16

! ! ! 8 on this a pproach, the researchers found that LBCS reported spirituality and/or religiosity as key coping strategies. Research has further highlighted the importance of SWB for oncology patients, including breast cancer patients and survivors, and especially for La tino cancer patients and survivors. Foundational research evidences SWB as a significant predictor of HRQOL among LBCS (Yanez et al., 2011). Brady et al. (1999) found that SWB has unique effects on both subjective and objective health outcomes and HRQOL am ong people who have had a serious illness(es) including breast cancer. Bredle et al., (2011) confirmed these findings. In addition, individuals who engage in organized religious activities tend to experience better psychological and physical health outcome s (Gallo et al., 2009). Moreover, Yanez et al. (2011) highlighted multiple studies that associated higher SWB with higher reports of HRQOL, and demonstrated that Latinas with breast cancer utilized spiritual and/or religious coping more frequently than NLW with breast cancer. As a result, SWB, similar to social support, may serve as a protective factor against adverse health outcomes for Latina breast cancer patients and survivors (Yanez et al. 2011). Perceived social support Among LBCS Social support is an important Latino cultural value. Social support may influence health outcomes depending on one's perception of how much social support he/she experiences. In general, perceived social support involves an individual's beliefs and experiences about how much he/she is cared for, esteemed, and valued, as well as the extent to which one perceives him /herself to be connected to a social network (e.g., family, friends, community) . One's social network might also involve people to whom one can turn for emotional and/or tangible (e.g., material, financial, etc.) social support (Crookes et al., 2016). There is a considerable body of research on the relationship between social support and QOL among cancer patients and

PAGE 17

! ! ! 9 survivors in general. Kroenke et al. (2013b) foun d that positive social interaction was significantly related to multiple measures of QOL including breast cancer specific QOL, HRQOL, physical wellbeing, social wellbeing, and emotional wellbeing. In previous work, Kroenke et al. (2013a; 2012) found that t he association between social networks and breast cancer outcomes was related to high and low levels of social support (i.e., emotional support, tangible support, affection, and positive interaction) and social burden (i.e., caregiving responsibilities, so cial strain or negative aspects of social relationships) within relationships. For example, low levels of support and high relationship burden were independently associated with higher mortality rates among women with cancer. Among LBCS, the perception of social support can influence QOL outcomes. In Yanez et al.'s (2011) systematic review, multiple studies found that perceived social support in general was positively correlated with QOL among LBCS. Graves et al. (2012) similarly found that perceived social support was significantly and positively associated with QOL. Additionally, among immigrant LBCS, those who reported less perceived social support tended to also report poorer QOL compared to other racial/ethnic groups. Immigrant LBCS often leave behind f amily and friends when they immigrate to the U.S. and may take time to build new social connections (Crookes et al., 2016). However, social support is not only an important predictor of QOL but it has been found to be an important predictor of how much ben efit cancer survivors may glean from the cancer experience (Weaver at al., 2008). Benefit Finding and Cultural Factors Among LBCS There is a substantial and growing body of research on the positive ways in which one's life may change after a traumatic eve nt or illness such as cancer. Researchers often label these positive changes as "post traumatic growth", "stress related growth", or "benefit finding" (BF)

PAGE 18

! ! ! 10 (Hegelson, Reynolds & Tomich, 2006). Although theories on BF vary slightly, they all agree that BF i nvolves the ability to find benefits and/or make positive life changes following a traumatic event (e.g., cancer diagnosis, treatment, etc.). BF may further be conceptualized as a coping strategy by which cancer patients and survivors may manage acute and chronic health issues and find meaning in the experience. Finding meaning in one's experience may be especially useful for those who face the long term effects of their cancer and treatment such that they may regain a perceived sense of mastery and control over their situation. In these ways, BF can influence physical and psychological health outcomes (Pascoe & Edvardsson, 2013). Previous research with BCS showed that higher perceived benefits were generally associated with better coping, higher QOL, and m ore positive overall health outcomes (Weaver et al., 2008; Tomich & Hegelson, 2004). Hegelson, Reynolds & Tomich (2006) conducted a meta analysis of 87 studies and found that BF was associated with less depression and greater positive wellbeing across dive rse racial/ethnic populations who experienced a variety of traumatic experiences (e.g., cancer, war stress, etc.). Tomich & Hegelson (2004) found that minority (i.e., Latinas and African American) women with breast cancer tended to perceive greater benefit s, ( t (360) = 3.01, p = .01) and to have less negative affect ( t (360) = 4.09, p = .01) compared to NLW with breast cancer. Researchers suggested that lower SES and greater disease severity led minority women to engage more in BF compared to NLW women. Rese archers also indicated that relationships among low SES, race/ethnicity, and BF might have been due to spiritual and/or religious coping strategies that involve cognitive restructuring of events. Additionally, individuals who face high adversity in general may be more apt to find benefit in negative events. However, Tomich & Hegelson (2004) also found that greater BF, regardless of race/ethnicity, was marginally associated with negative affect ( r = .09, p = .10).

PAGE 19

! ! ! 11 Given all this, research findings suggest t hat BF might be both an important coping strategy and psychological health outcome for such minority groups as LBCS. However, more research is needed to elucidate which factors or mechanisms predict BF and HRQOL among LBCS (Pascoe & Edvardsson, 2013; Garci a Jimenez, 2013; Graves et al., 2012; Martinez Ramos et al., 2013; Yanez et al., 2011). In particular, a better understanding of how spirituality and social support may function as protective factors to improve BF and HRQOL in LBCS is necessary . Such data can then be used to highlight and explain the risks that LBCS face, as well as new ways in which cancer survivorship programs may be improved for growing Latino communities in the U.S. In this way, future generations of Latinos may face fewer challenges an d threats to overall physical and psychological health. The Present Study The purpose of this study was to explore the relationships among spiritual wellbeing, perceived tangible social support, health related quality of life, and benefit finding in a samp le of Latina breast cancer survivors (LBCS) from Arizona and Colorado. The follow up study SUNSHINE is a population based, case control study that has allowed researchers to evaluate the differences between LBCS and Non Latino White breast cancer survivors (NLWBCS) with regard to breast cancer risk based on genetic and lifestyle factors. The sample population was also at least 6 years post treatment, such that they may be considered long term cancer survivors. According to the American Cancer Society, a lon g term cancer survivor is at least 5 years post treatment (Chopra & Kamal, 2012). SUNSHINE utilized reliable markers of sociodemographic variables (e.g., age, marital status), medical variables (e.g., tumor stage at diagnosis, comorbidities), as well as su ch psychosocial variables as quality of life, spiritual wellbeing, perceived tangible social support, and benefit finding (Risendal et al., 2014) . Therefore,

PAGE 20

! ! ! 12 SUNSHINE offers an excellent opportunity to explore the relationships among such psychosocial vari ables among LBCS while controlling for differences on confounding variables. A priori power analyses were conducted to determine whether or not the proposed sample sizes from the SUNSHINE project would be large enough to detect a medium effect size using the proposed analyses. The first power analysis revealed that a total sample size of at least 128 is required to detect a medium effect size ( f = .25) using the proposed analytic approach to test the preliminary hypotheses under Aim 1. Thus, the total samp le of BCS cases ( N = 360) will be sufficient to achieve enough statistical power to detect a medium effect size. The second power analysis revealed that a total sample size of 55 LBCS cases would be required to achieve enough statistical power to detect a medium effect size ( f = .15) using the proposed analytic approach to test the proposed hypotheses under Aims 2 4. Thus, the total sub sample of LBCS cases ( n = 105) was proposed to be sufficient to achieve enough statistical power to detect a medium effect size. Hypotheses To evaluate the independent effects of perceived spiritual wellbeing and perceived tangible social support on health related quality of life and benefit finding in the survivorship period for LBCS, the following aims and hypotheses wi ll be evaluated: Aim 1) To examine whether there are differences between LBCS and NLWBCS on measures of spiritual wellbeing, perceived tangible social support, HRQOL, and BF when controlling for sociodemographic (i.e., age, education level, marital status) and medical variables (i.e., tumor stage at diagnosis, comorbidities). Hypothesis 1a : LBCS will report higher spiritual wellbeing than NLWBCS when controlling for sociodemographic and medical variables.

PAGE 21

! ! ! 13 Hypothesis 1b : LBCS will report higher perceived ta ngible social support than NLWBCS when controlling for sociodemographic and medical variables. Hypothesis 1c: LBCS will report lower HRQOL than NLWBCS when controlling for sociodemographic and medical variables. Hypothesis 1d: LBCS will report higher benef it finding than NLWBCS when controlling for sociodemographic and medical variables. Aim 2) To examine whether spiritual wellbeing is a significant predictor of HRQOL and benefit finding in LBCS. Hypothesis 2a : Spiritual wellbeing will be a significant pre dictor of HRQOL in LBCS when controlling for sociodemographic and medical variables, such that higher levels of self reported spiritual wellbeing predicts better HRQOL. Hypothesis 2b : Spiritual wellbeing will be a significant predictor of benefit finding in LBCS when controlling for sociodemographic and medical variables, such that higher levels of spiritual wellbeing predict more benefit finding. Aim 3) To examine whether perceived tangible social support is a significant predictor of HRQOL and benefit fi nding in LBCS. Hypothesis 3a: Perceived tangible social support will be a significant predictor of HRQOL in LBCS when controlling for sociodemographic and medical variables, such that higher levels of perceived tangible social support predict better HRQOL . Hypothesis 3b : Perceived tangible social support will be a significant predictor of benefit finding in LBCS when controlling for sociodemographic and medical variables, such that higher levels of perceived tangible social support predict more benefi t finding. Exploratory Aim 4) Examine whether benefit finding predicts HRQOL in LBCS.

PAGE 22

! ! ! 14 Hypothesis 4: Benefit finding will predict HRQOL in LBCS when controlling for sociodemographic and medical variables such that more benefit finding predicts better HRQOL .

PAGE 23

! ! ! 15 C hapter II Methods Sample The Survivorship Update Network to Southwest Hormone, Insulin, Nutrition, and Exercise (SUNSHINE) ( N = 725) is a population based, case control study that has allowed researchers to evaluate the differences between Latina breas t cancer survivors (LBCS) and Non Latina breast cancer survivors (NLWBCS) from Arizona and Colorado with regard to breast cancer risk based on lifestyle and genetic factors (Risendal et al., 2014) . SUNSHINE took place in Arizona and Colorado between April 2007 and July 2008. Although 1,969 women completed the parent study [i.e., Southwest Hormone, Insulin, Nutrition, and Exercise (SHINE) 4 Corners Breast Cancer Study] , those who met the following criteria were considered ineligible for the follow up study S UNSHINE: 1) they were deceased; 2) they had a mental or physical disability that would inhibit them from completing the interview; 3) they were living in a nursing home or group facility. There were 1,969 women who completed the study. Of the 1,969 women w ho completed the SHINE study, about 1,572 women were considered e ligible for SUNSHINE. A bout 1,068 consented to participate in SUNSHINE . However, data was considered incomplete if participants completed either the written interview only ( n = 70) or the tel ephone interview only ( n = 150). Data was also considered incomplete if participants provided incomplete surveys ( n = 123). Given all this, about 725 participants were found to have provided complete data. Because the present study aimed to evaluate the di fferences between LBCS and NLWBCS, all control studies ( N = 365) from SUNSHINE were excluded. Case studies ( n = 360) were categorized into LBCS ( n = 105). Additionally, women were excluded from the analyses if they reported a recurrence and/or more than 1 occurrence of breast cancer. In these ways, the

PAGE 24

! ! ! 16 analyses might better shed light on the differences between LBCS and NLWBCS on variables known to be important in BCS populations ( Aim 1). The analyses might also shed light on variables that may serve as key predictors of HRQOL and BF in LBCS ( Aims 2 4). Procedure Research assistants at the Research Core, University of Colorado Cancer Center (Aurora, CO), screened and recruited participants from the SHINE 4 Corners Breast Cancer Study. Descriptions of study design and recruitment methods of this parent study have been published (Risendal et al., 2015; Sedjo et al., 2013). Between April 2007 and July 2008, research assistants randomly selected and contacted 20 40 potential participants each week. The study em ployed the Dillman protocol for recruitment, which involved the following steps: 1) Research assistants mailed an introductory letter and project brochure; 2) Two weeks later, research assistants mailed a packet that contained the introductory letter, proj ect brochure, and a written questionnaire; 3) If participants did not respond within four to six weeks, research assistants sent reminder postcards to the participants (Risendal et al., 2014; Hoddnott & Bass, 1986). Research assistants then scheduled telep hone interviews with consenting participants. Data that was not collected via written surveys was again collected via telephone interviews in addition to other questionnaires. The questionnaires allowed researchers to collect a variety of sociodemographic, medical, and psychosocial data ( Risendal et al., 2014). As mentioned above, study analyses first focused on all breast cancer cases ( N = 360) ( Aim 1), and then focused solely on LBCS ( n = 105) ( Aims 2 4). Measures Spiritual wellbeing Spiritual wellbeing was assessed with the Functional Assessment of Chronic Illness Therapy Ð Spiritual Wellbeing Scale version 4 (FACIT Sp 12) and was available in English and

PAGE 25

! ! ! 17 Spanish. The FACIT Sp 12 is a unidimensional, quantitative measure that has been widely used and va lidated with cancer populations. The FACIT Sp 12 is a 12 item self report measure that assesses two aspects of spiritual wellbeing: Meaning/Peace and Faith. Items were meant to address a respondent's sense of meaning in life, harmony, peacefulness, and to what extent the respondent is able to derive strength and comfort from his/her faith. The FACIT Sp 12 utilizes 5 point Likert style responses anchored as: (0) " Not at all "; (1) " A little bit "; (2) " Somewhat "; (3) " Quite a bit "; (4) " Very much ." Responses m ay be summed, yielding a score between 0 and 48 with higher scores indicating higher spiritual wellbeing. The FACIT Sp 12 was found to have high internal consistency ( Cronbach's ! = .87 ) . The subscales were found to be moderately and positively correlated with the FACT G ( r s = .58, p < .001). However, the Meaning/Peace subscale was more strongly (positively) correlated with the FACT G ( r s = .62, p < .001). The FACIT Sp 12 also dem onstrated substantial discriminant validity such that there was an inverse relationship between FACIT Sp 12 scores and POMS depression subscale scores [ F (2, 1586) = 186.98, p = .0001]. Perceived tangible social support Perceived tangible social support wa s assessed using the tangible social support subscale from the 40 item Interpersonal Support Evaluation List (ISEL). The perceived tangible social support subscale includes 10 items and has been used to assess one's perception of the material or financial aid available to him/h er, as well as support with daily chores or activities (Uchino, 2004) (e.g., "If I needed help fixing an appliance or repairing my car, there is someone who would help me," "It would me difficult to find someone who would lend me thei r car for a few hours," "If I had to go out of town for a few weeks, it would be difficult to find someone who would look after my house or apartment (the plants, pets, garden, etc.)." The ISEL and

PAGE 26

! ! ! 18 subsequent versions (e.g., ISEL Ð Short Form, ISEL 12) hav e been widely accepted and utilized in psychosocial research, especially as a predictor of psychological and physical health indices (Uchino, 2004). The ISEL overall demonstrates good internal consistency (Cronbach's ! = .68). More specifically, the tangib le social support subscale demonstrated good test retest reliability (.69) within a student population . The perceived tangible social support subscale is available in English and Spanish. The tangible social support subscale utilizes 4 point Likert style r esponses anchored as: (0) "Definitely true"; (1) "Probably True"; (2) "Probably false"; (3) "Definitely false" (Cohen & Hoberman, 1983). Research has demonstrated the psychometric validity and reliability of the ISEL as a general population measure (Brook ings & Bolton, 1988). Brookings & Bolton (1988) conducted confirmatory factor analysis (CFA) of the ISEL with 133 college students who participated in a longitudinal study of stress and depression. The CFA demonstrated that the ISEL has substantial structu ral validity such that there are 4 factors of perceived social support, including tangible, belonging, appraisal, and self esteem (Brookings & Bolton, 1988). The ISEL was also found to have high convergent validity with other established measures of social support, as well as sufficient discriminant validity with personality measures (Cohen, Mermelstein, Kamarck, & Hoberman, 1985, as cited in Brookings & Bolton, 1988). HRQOL HRQOL was assessed with the Functional Assessment of Cancer Therapy Scale (FACT G) version 4 and was available in English and Spanish. The FACT G version 4 is a 27 item self report measure, in which items evaluate four dimensions of wellbeing: physical (PWB), functional (FWB), social/family (SFWB), and emotional (EWB). The PWB subscale refers to physical symptoms that the respondent may have experienced. The FWB refers to the extent to

PAGE 27

! ! ! 19 which a respondent is able to participate in and enjoy everyday activities. The SFWB subscale refers to the types of social support one might receive from family and friends. The SFWB subscale includes 7 items that has been used to assess one's perception of closeness or emotional support received from family and friends (e.g., "I feel close to my friends," "I get emotional support from my family") . In this way, SFWB subscale addresses different aspects of social experience compared to the tangible social support subscale of the ISEL. The EWB subscale refers to the respondent's general mood and emotional response to the illness (Ashing Giwa & Rosales, 2013). All of the subscales utilize 5 point Likert style responses anchored as: (0) "Not at all"; (1) "A little bit"; (2) "Somewhat"; (3) "Quite a bit"; and (4) "Very much." To score the FACT G, participants' responses may be summed, yielding a score between 0 a nd 1 12 . Higher scores indicate higher HRQOL (Cella et al., 1993) . The FACT G is a unidimensional, quantitative measure that is considered one of the most reliable and valid measures of HRQOL in chronic illness populations including cancer (Ashing Giwa & R osales, 2013). In terms of reliability, analyses revealed an overall Cronbach's alpha score of .90 on the English version. The subscales had similarly high alpha coefficient scores (PWB, .82; FWB, .80; SFWB, .69; EWB, .74). In terms of validity, the FACT G has shown substantial convergent validity such that it was significantly correlated ( r = .79, p < .05) with another well established QOL measure, the Functional Living Index Cancer measure, as well as with the well established mood distress scale Taylor M anifest Anxiety Scale ( r = .58, p < .05) and Brief Profile of Mood States (POMS)( r = 0.65, p < 0.05). The FACT G has also shown substantial divergent validity such that it has been found to have a low correlation with the Marlowe Crowne Social Desirabili ty Scale ( r = .22, p < .05). In addition, the FACT G has been

PAGE 28

! ! ! 20 found to be highly sensitive to disease stage, especially on the PWB and FWB subscales (Cella et al., 1993). Moreover, the FACT G Spanish version has been found to have high concurrent validity with the English version of the FACT G, as well as high content and semantic validity (Cella et al., 1998). The FACT G Spanish version has also been more specifically cross validated with Latina breast cancer populations. The FACT G demonstrated high cons truct validity such that the four factors accounted for 56 percent of the variance in HRQOL among English language proficient Latinas and 57 percent among limited English language proficient Latinas. The FACT G also demonstrated high internal consistency ( Cronbach's ! = .91 ) among English language proficient and limited English language proficient Latinas with breast cancer (Ashing Giwa & Rosales, 2013). Perceived benefit of the breast cancer experience Perceived benefit or benefit finding (BF) was assessed using th e Benefit Finding (BF) Scale for breast cancer, which was available in English and Spanish. The BF scale is a 17 item self report measure that assesses the benefits that a respondent may perceive as a result of diagnosis and treatment of breast cancer. The BF scale is a unidimensional, quantitative measure that has been commonly used and validated in cancer populations. Subscales include: acceptance, sensitivity to others, improved coping, and new purpose of life. The BF scale utilizes 5 point Likert style responses anchored as: (0) "Not at all"; (1) "A little"; (2) "Moderately"; (3) "Quite a bit"; (4) "Extremely." To score the BF scale, participants' responses may be summed, yielding a score between 0 and 85. Higher scores indicate higher perceived benefit. The BF scale was found to have high internal consistency (Cronbach's ! = .95 ) . The BF scale was also found to have convergent and discriminant validity. It was somewhat positively correlated with

PAGE 29

! ! ! 21 optimism ( r = .23), and inversely (but again not overwhelmingly) correlated with a POMS derived index of distress ( r = .25) (Boy er et al., 2000 in Antoni et al., 2001). Race/Ethnicity Race/ethnicity was measured as a single item in order to observe differences between NLW and Latina breast cancer survivors. As mentioned above, previous literature ( Ashing Giwa et al., 2007; Maly e t al., 2008; Yanez et al., 2011) , demonstrates significant differences among racial/ethnic groups on survivorship outcomes, including NLW and Latina breast cancer survivors. Covariates Covariates were included in order to increase statistical power and re duce the probability of a Type II error. The covariates assessed in the present study included sociodemographic and medical variables. These covariates were chosen based on previous analyses from the same dataset (Risendal et al., 2015; Sedjo et al., 2013) , as well as previous literature on the topic of breast cancer survivorship in multiethnic samples including LBCS, (Garcia Jimenez et al., 2013; Janz et al., 2014; Maly et al., 2008), which suggest that the covariates chosen for this project might account for some of the variance in such outcomes as HRQOL and BF in breast cancer survivors, including LBCS. Sociodemographic variables . The sociodemographic variables that were included in the analysis of covariance (ANCOVA) analyses were race/ethnicity, age, education level, and marital status. The sociodemographic variables that were included in the multiple regression models did not include race/ethnicity, and only included the following sociodemographic variables: age, education level, and marital status.

PAGE 30

! ! ! 22 Age at the time participants were interviewed was measured as a single item and was coded as a continuous variable. Education level was measured as a single item and was coded as a categorical variable with 4 categories, which included the following: 1 = < High School; 2 = High School; 3 = Some college; and 4 = College +. Marital status was measured as a single item and was coded as a categorical variable with 6 categories, which included the following: 1 = Married; 2 = Living as married; 3 = Widowed; 4 = Divorced; 5 = Separated; and 6 = Never married. In order to account for the differences between having a long term partner (defined as married or living as married) and being single, the variable was re coded and entered into each multiple regression anal ysis as a binary, dummy variable as follows: 1 = Married or Living as married; and 2 = Currently single (widowed, divorced, separated, never married). Medical variables . The medical variables that were included in the multiple regression models were tumo r stage at diagnosis and the number of self reported comorbidities. Tumor stage at diagnosis was measured using the Surveillance Epidemiology and End Result (SEER) cancer registry or the state tumor registry database. The second edition of the Internationa l Classification of Diseases for Oncology (ICD O 2) was used for breast cancer cases diagnosed prior to 2001, whereas the ICD O 3 was used to code cases diagnosed after 2001. The histologic types were categorized as follows: ductal carcinoma (8230, 8500, 8521, 8523), lobular carcinoma (8520, 8524), ductal/lobular (8522), all others, and unknown. Tumor stage classifications were then determined according to SEER summary stage codes using the 1977 definitions for breast cancer cases diagnosed prior to 2001. For breast cancer cases diagnosed after 2001, the 2000 definitions were used. The tumor stage at diagnosis variable was then coded as a categorical variable as follows: 0 = In situ (i.e., pathological stage 0); 1 = Local; 2 = Regional by direct extension ( D.E.); 3 = Regional lymph nodes only involved; 4 = Regional by

PAGE 31

! ! ! 23 both D.E. and to regional nodes; 5 = Regional not otherwise specified (NOS); 7 = Distant sites and/or distant nodes; 8 = CNS (benign or borderline); and 9 = Unknown or Not applicable. However, previous analyses of the dataset condensed the variables into the following: 0 = In situ; 1 = Localized; 2 = Regional/Distant; 3 = Unstaged or Missing (Sedjo et al., 2013). Therefore, for analyses in the present study, and to account for the differences be tween non metastasized and metastasized cancer, the tumor stage at diagnosis variable was entered into the statistical models as a binary, dummy variable as follows: 1 = In situ or localized; and 2 = Regional or regional NOS or distant. Comorbidities were measured using the Charlson Comorbidity Index (CCI). The CCI allows researchers to categorize participants' comorbidities according to the International Classification of Diseases 9. The CCI provides a weight for each comorbidity based on the adjusted risk of mortality or the amount of healthcare resources utilized by a patient. The weights are summed to produce a single comorbidity score for each participant. Higher CCI scores are associated with an increase in the adjusted risk of mortality or higher reso urce use (University of Manitoba, 2016). The CCI variable used in the present study was measured categorically between 0 and 17 comorbidity categories. Because not many participants had 3 or more comorbidities, it was not useful to maintain 17 categories. Most participants were distributed across 3 levels of categories; therefore, this variable was entered into the statistical models as a tertile variable as follows: 0 = 0; 1 = 1 or 2; 3 = 3 to 14. Statistical Analysis Before conducting the main analysis, data screening was conducted and assumptions for ANCOVA and hierarchical multiple regression analyses were tested. In order to conduct such multivariate analyses, IBM SPSS Statistics for Windows, version 24 (IBM Corp., Armonk, N.Y.,

PAGE 32

! ! ! 24 USA) was used. The sign ificance level for all statistical analyses was ! < .05 (two tailed). Descriptive statistics of the Whole Sample ( N = 360), LBCS group ( n = 105) , and the NLWBCS ( n = 255) , respectively, were subsequently computed, which included means, standard deviations, frequencies, and percentages (see Tables 1 and 2). Next, bivariate correlations between all pairs of variables of the Whole Sample ( N = 360) and the LBCS group ( n = 105), respectively, were conducted (see Tables 4 and 5). These bivariate correlations allo wed for the interpretation of direction and strength of any potential correlations between each pair of variables. Based on the bivariate correlation models between predictor and outcome variables, multicollinearity was not observed among the hypothesized covariates (i.e., age, education level, marital status, tumor stage at diagnosis, and comorbidities ) . However, t he hypothesized covariates were correlated with the primary variables of interest; therefore, they were maintained across all analyses . ANCOVA s were then conducted to evaluate Aim 1 or group mean differences on the primary variables (i.e., SWB, PTSS, HRQOL, and BF ) . In order to evaluate Aims 2 4, h ierarchical multiple linear regression s were conducted , because hierarchical multiple linear regressi on allow s researchers to specify a fixed order of entry for variables, and as a result, control for the effects of specific covariates. The m odel s generally included sociodemographic variables in the first block, medical variables in the second block, and the respective predictor variable in the third block. Overall, the statistical models illustrated the extent to which each predictor variable w as independently associated with respective outcome variables above and beyond the effects of the sociodemographi c and medical variables.

PAGE 33

! ! ! 25 Chapter III Results Demographics and Descriptive Statistics in the Whole Sample The total sample of breast cancer survivor cases ( N = 360) included more NLWBCS (71%) than LBCS (29%). Participants' ages ranged from 33 to 85 ( M = 61.37; SD = 9.9 9 ). Most participants completed at least some college (73%) and were married or living as married (74%). More participants were initially diagnosed with in situ or local cancer (68%) compared to regional, regional NOS, or distant cancer (26% ). With regard to comorbidities, 52% of all breast cancer survivors reported a Charlson Comorbidity Index (CCI) score of 0, 27% had a CCI score of 1 or 2, and 19% had a CCI score of 3 Ð 14. Table 1 (below) provides an overview of the demographic informati on for the Whole Sample. Table 2 (below) provides an overview of the descriptive information for the Whole Sample. Table 1 . Demographic Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), & NLWBCS Group (n = 255) Participant Characteristic Min imum Maximum M (SD) n % Age Whole Sample 33 85 61.37 (9.99) 360 LBCS 33 83 59.19(10.64) 105 NLWBCS 43 85 62.27(9.59) 255 Race/Ethnicity Latina 105 29.2 Non Latina White 255 70.8 Education l evel Whole Sample < HS 34 9.4 HS 65 18.1 Some college 130 36.1 College+ 131 36.4 LBCS
PAGE 34

! ! ! 26 Table 1 cont'd College+ 20 19 .0 NLWBCS
PAGE 35

! ! ! 27 Table 1 cont'd 1 or 2 96 26.7 3 14 69 19.2 LBCS 0 47 44.8 1 or 2 26 24.8 3 14 32 30.5 NLWBCS 0 148 58 1 or 2 70 27.5 3 Ð 14 37 14.5 Note . HS = high school; NOS = not otherwise spe cified; CCI = Charlson comorbidity index Table 2. Descriptive Statistics for the Whole Sample (N = 360), LBCS Group (n = 105), & NLWBCS Group (n = 255) Participant Characteristic Minimum Maximum M (SD) N SWB Whole Sample 5 48 37.96(8. 36) 356 LBCS 13 48 38.84( 8.09) 104 NLWBCS 4 48 37.60(8.45) 252 PTSS Whole Sample 10 38 23.94(3.18) 353 LBCS 10 38 23.45(4.13) 104 NLWBCS 10 37 24.15(2.67) 249 HRQOL Whole Sample 34.40 112 92.74(14. 80) 348 LBCS 34.40 112 88.59(16.70) 102 NLWBCS 49 112 94.46(13.62) 246 BF Whole Sample 17 85 59.56 (14.94) 342 LBCS 30 85 64.21(12.611) 101 NLWBCS 17 85 57.61(15.43) 241 Note . SWB = spiritual wellbeing; PTSS = perc eived tangible social support; BF = benefit finding; HRQOL = health related quality of life. Demographics and Descriptive Statistics in the LBCS Group

PAGE 36

! ! ! 28 In the Latina breast cancer survivors (LBCS) group ( n = 105), participants' ages ranged from 33 to 83 ( M = 59.19; SD = 10.6 4 ). Participants' education level varied across the four categories, su ch that 25.7 % had less than a high school education, 21% graduated high school, 34.3 % ha d some college education, and 19 % had at least a college degree. Most partici pants were married or living as married (71%). More participants were originally diagnosed with in situ or cancer (65%) compared to regional, regional NOS, or distant cancer (28%). With regard to comorbidities, about 45% of the LBCS had a Charlson Comorbid ity Index (CCI) score of 0. About 25% had a CCI score of 1 or 2, and about 31% had a CCI score of 3 Ð 14. Table 1 ( above ) provides an overview of the demographic information for the LBCS group. Table 2 (above) provides an overview of the descriptive infor mation for the LBCS group. Demographics and Descriptive Statistics in the NLWBCS Group In the NLWBCS group ( n = 255), participants' ages ranged from 43 to 85 ( M = 62.27; SD = 9.59 ). Participants' education levels were such that 2.7% had less than a high school education, 16.9% graduated high school, 36.9% had some college education, and 43.5% had at least a college degree. Most participants were married or living as married (76.1%). More participants were originally diagnosed with in situ or cancer (69.8 %) compared to regional, regional NOS, or distant cancer (24.7%). With regard to comorbidities, about 58% of the NLWBCS had a CCI score of 0. About 27.5% had a CCI score of 1 or 2, and about 14.5% had a CCI score of 3 Ð 14. Table 1 ( above ) provides an ove rview of the demographic information for the NLWBCS group. Table 2 (above) provides an overview of the descriptive information for the NLWBCS group. Assumptions

PAGE 37

! ! ! 29 Analyses involved the Whole Sample of breast cancer survivors ( N = 360), which included LBCS g roup ( n = 105) and NLWBCS group ( n = 255). The evaluation of model assumptions for ANCOVA and hierarchical multiple linear regression revealed that there was no multicollinearity (i.e., correlation above . 8 ) among any of the variables, including predictor, outcome, and covariate variables. Analysis of residuals included the calculation of Mahalanobis distance, which revealed a few outliers in the data; however, the outliers did not have a major impact on the data . The sample sizes were large enough and para meter estimation involved robust standard errors. There did not appear to be any patterns in the data with regard to multivariate outliers and missing data , and no variable exhibited more than 5 percent of missing cases. Assumptions for homoscedasticity (f or the hierarchical multiple linear regression analyses) and homogeneity of variance (for the ANCOVA analyses) were not always met ; however, these findings did not impact the respective analyses . Additionally, heterogeneity of the variances might reflect d ifferences in the data based on race/ethnicity, which is a variable observed specifically in the ANCOVA analyses. Thus, in order to maintain the integrity of the data, transformations were not conducted. Bivariate Correlation Analyses with the Whole Sampl e Pearson correlation analyses were run among the covariates and primary variables in the Whole Sample . A s described above, the analyses did not reveal any multicollinearity. However, the analyses did reveal several significant relationships. Race/ethnici ty (1 = NLW; 2 = Latina) was negatively associated with age ( p < 0.01 ) , education ( p < .001 ) , and health related quality of life (HRQOL) ( p < 0.01 ) , indicating that LBCS were lower then NLWBCS on all three of these variables. Race/ethnicity was positivel y correlated with comorbidities ( p < .01 ) and benefit finding (BF ; p < .001 ) . Age was positively correlated with marital status ( p < .01 ) , comorbidities

PAGE 38

! ! ! 30 ( p < .001 ) , and spiritual wellbeing (SWB ; p < .05 ) . Education was negatively correlated with comorbidit ies ( p < .001 ) and BF ( p < .01 ) . Education was positively correlated with HRQOL ( p < .05). Marital status was negatively correlated with tumor stage at diagnosis ( p < .01 ) , perceived tangible social support (PTSS) ( p < .001 ) , and HRQOL ( p < .01 ) . Comorbidi ties were negatively correlated with HRQOL ( p < .001 ) . Among the primary variables, SWB was positively correlated with HRQOL ( p < .001 ) and BF ( p < .001 ) . BF was positively correlated with HRQOL ( p < .01 ) . Results of the bivariate correlations among covari ates and primary variables in the Whole Sample ( N = 360) are presented in Table 3 below . Table 3 Bivariate Correlations of Covariates and Primary Variables in the Whole Sample (N = 360) 1 2 3 4 5 6 7 8 9 10 1. Race/ Ethnicity _ _ .14** .35** .06 .04 .17 . 07 .10 .20 .18** 2. Age __ .07 .14 ** .00 .21** .12 * .08 .01 .09 3. Educatio n __ .02 .05 .19** .01 .09 .16** .11* 4. Marital Status __ .14** .04 .06 .20** .00 .17** 5. Tumor stage at diagnosis __ .01 .02 .03 .03 .00 6. CCI score __ . 06 .10 .02 .26** 7. SWB __ .01 .33** .67** 8. PTSS __ .00 .10 9. BF __ .14** 10. HRQOL __ Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; PTSS = perceived tangible social support; BF = benefit finding; HRQOL = he alth related quality of life; * p < .05; ** p < .01 Bivariate Correlations with the LBCS Group The bivariate correlations with the LBCS group revealed several significant relationships . Age was positively correlated with marital status ( p < .05 ) and comorb idities ( p < .05 ) . Age was

PAGE 39

! ! ! 31 negatively correlated with education ( p < .05 ) . Marital status was negatively associated with HRQOL ( p < .01 ) . Comorbidities were negatively correlated with HRQOL ( p < .001 ) . SWB was positively correlated with BF ( p < .001 ) and HRQOL ( p < .001 ) . Results of the bivariate correlations among covariates and primary variables with in the LBCS group ( n = 105) are presented in Table 4 . Table 4 Bivariate Correlations of Covariates and Primary Variables in the LBCS Group (n = 105) 1 2 3 4 5 6 7 8 9 1. Age __ .23* .22* .06 .21* .12 .10 .15 .16 2. Education __ .01 .10 .18 .03 .12 .11 .02 3. Marital Status __ .08 .06 .11 .18 .02 .26** 4. Tumor stage at diagnosis __ .04 .02 .04 .05 .02 5. CCI score __ .13 .10 .10 .40** 6. SW B __ .04 .34 * .69** 7. PTSS __ .01 .07 8. BF __ .16 9. HRQOL __ Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; PTSS = perceived tangible social support; BF = benefit finding; HRQOL = health related quality of life ; * p < .05; ** p < .01 Group Differences on Covariates Chi squared analyses demonstrate d significant group differences on education [ " 2 (3) = 54.623, p < .001] and comorbidities (i.e., CCI scores) [" 2 (2) = 12.514, p = .01] , such that NLWBCS had significantly more education and fewer comorbidities compared to LCBCS . However, there were no drastically large differences between the LBCS and NLWBCS groups on the covariates ( see Tables 5 and 6 below); therefore, all 5 covariates were maintained across the main analyses (i.e., ANCOVA and hierarchical regression models).

PAGE 40

! ! ! 32 T able 5 One Way Analysis of Variance (ANOVA) of Racial/Ethnic Differenc es on the Continuous Covariate Age Variable df F Effect Size (partial ! 2 ) Age 1 7.21** .02 Note . ** p < .01 T able 6 Tests of Racial/Ethnic Differences on Categorical Covariates Race/Ethnicity Variables LBCS ( n ) NLWBCS ( n ) " 2 D f E ducation 105 255 54.62*** 3 Marital Status 105 255 1.23 1 Tumor stage at diagnosis 105 255 1.08 2 CCI score 105 255 12.51** 2 Note . LBCS = Latina breast cancer survivors; NLWBCS = Non Latina White breast cancer survivors; * p < .05; ** p < .01; ***p < .001. Group Differences on Primary Variables (Aim 1) ANCOVA analyses for Aim 1 tested whether there were group differences on the 4 primary variables of interest when controlling for all 5 covariates (i.e., age, education, marital status, tumor stage a t diagnosis, and comorbidities). Results showed that LBCS had higher mean scores on the SWB [ F (1, 356)= 4.99 , p < .05] and BF [ F (1, 342) = 8.77 , p < .01] compared to the NLWBCS. The ANCOVA analyses showed no significant differences between the LBCS and N LWBCS mean scores on PTSS [ F (1, 353) = 1.54, ns ] or HRQOL [ F (1, 348) = 2.96, ns ] (see Table 7 on page 33 ) .

PAGE 41

! ! ! 33 T able 7 One Way Analysis of Covariance (ANCOVA) of Racial/Ethnic Differences on Primary Variables with Sociodemographic and Medical Covaria tes Variable Df F Effect Size (partial ! 2 ) SWB 1 4.99* .01 PTSS 1 1.54 .00 HRQOL 1 2.96 .01 BF 1 8.77** .03 Note . SWB = spiritual wellbeing; PTSS = perceived tangible social support; HRQOL = health related quality of life; BF = benefit finding; * p < .05; ** p < .01. Hierarchical Multiple Regression Analyses ( Aims 2 3) with the LBCS Group Hierarchical multiple regression analyses for Aims 2 3 examined hypothesized effects within the LBCS group on the primary variables and controlling for covariates . S WB was positively associated with HRQOL when controlling for all 5 covariates [ F (6, 94) = 23.551, p < .001] . SWB was, as hypothesized, a strong predictor of HRQOL in the LBCS [ ! = .640, p < .001). M arital status and comorbidities were additionally found t o be independent predictors of HRQOL [ ! = .177, p < .05 and ! = .289, p < .001, respectively] in the LBCS (see Table 8 below) . Table 8 Results of Hierarchical Multiple Regression for SWB as a Predictor of HRQOL when controlling for Sociodemographic an d Medical Covariates in the LBCS Group (n = 105) Variables B SE(B) # R 2 $ R 2 F .60 .39 23.55*** Age .03 .11 .02 Education .31 1.09 .02 Marital Status 6.46 2.47 .18* Tumor stage at diagnosis .00 .00 .03 CCI score 5.60 1.31 .29*** SWB 1.32 .14 .64*** Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; HRQOL = health related quality of life; Degrees of freedom of F = (6, 94); * p < .05; *** p < .001.

PAGE 42

! ! ! 34 Similarly, SWB was positively associated with BF when controlli ng for all 5 covariates [ F (6, 92) = 2.828, p < .05] . SWB was, as hypothesized, a strong predictor of BF in the LBCS group [ ! = .316, p < .01]. However, none of the covariates independently predicted BF (see Table 9 below ). Table 9 Results of Hierarchical Multiple Regression for SWB as a Predictor of BF when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105) Variables B SE(B) # R 2 $ R 2 F .16 .10 2.83* Age .18 .12 .15 Education 1.61 1.22 .13 Marital Status 2. 17 2.77 .08 Tumor stage at diagnosis .00 .00 .06 CCI score .86 1.48 .06 SWB .50 .16 .32** Note. CCI = Charlson comorbidity index; SWB = spiritual wellbeing; BF = benefit finding; Degrees of freedom of F = (6, 92); * p < .05; ** p < .01. T he analyses further showed that PTSS was positively associated with HRQOL when controlling for all 5 covariates [ F (6, 94) = 4.161, p < .01] . However, PTSS was not an independent predict or of HRQOL in the LBCS group ( ! = .026, ns ). Of the covariates, mari tal status and comorbidities independently predict ed HRQOL, [ ! = 2.267, p < .05 and ! = 3.983, p < .001, respectively] (see Table 1 0 below ). Table 1 0 Results of Hierarchical Multiple Regression for PTSS as a Predictor of HRQOL when controlling for Socio demographic and Medical Covariates in the LBCS Group (n = 105) Variables B SE(B) # R 2 $ R 2 F .21 .00 4.16** Age .07 .15 .05 Education .88 1.54 .06 Marital Status 7.97 3.51 .22* Tumor stage at diagnosis .00 .00 .02 CCI score 7.31 1.84 .38*** PTSS .01 .40 .00

PAGE 43

! ! ! 35 Table 10 cont'd Note. CCI = Charlson comorbidity index; PTSS = perceived tangible social support; HRQOL = health related quality of life; Degrees of freedom of F = (6, 94); * p < .05; *** p < .001. Analyses also r evealed that PTSS was not associated with BF when controlling for all 5 covariates [ F (6, 92) = .971, ns ]. Similarly n either the primary variable nor any of the covariates independently predict ed BF in the LBCS group (see Table 1 1 below ). Table 1 1 Results of Hierarchical Multiple Regression for PTSS as a Predictor of BF when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105) Variables B SE(B) # R 2 $ R 2 F .06 .00 .97 Age .22 .13 .18 Education 2.04 1.28 .17 M arital Status 1.95 2.96 .07 Tumor stage at diagnosis .00 .00 .06 CCI score 1.47 1.55 .10 PTSS .01 .32 .00 Note. CCI = Charlson comorbidity index; PTSS = perceived tangible social support; BF = benefit finding; Degrees of freedom of F = (6, 92); All F statistics were non significant; * p < .05; ** p < .01; *** p < .001. A nalyses additionally showed that BF was positively associated with HRQOL [ F = 3.509, p < .01) when controlling for all 5 covariates . However, BF was not a significant pre dictor of HRQOL ( ! = 1.220, ns ). Among the covariates, comorbidities were a significant predictor of HRQOL ( ! = 3.577, p < .01) (see Table 1 2 below ). Table 1 2 Results of Hierarchical Multiple Regression for BF as a Predictor of HRQOL when controlling for Sociodemographic and Medical Covariates in the LBCS Group (n = 105) Variables B SE(B) # R 2 $ R 2 F .19 .01 3.51** Age .03 .15 .02 Education .66 1.55 .04 Marital Status 6.84 3.51 .19 Tumor stage at diagnosis .00 .00 .02

PAGE 44

! ! ! 36 Table 12 c ont'd CCI score 6.64 1.86 .35** BF .15 .13 .12 Note. CCI = Charlson comorbidity index; BF = benefit finding; HRQOL = health related quality of life; Degrees of freedom of F = (6, 91); ** p < .01; *** p < .001

PAGE 45

! ! ! 37 Chapter IV Discussion Study fin dings demonstrated significant differences between LBCS and NLWBCS , which were consistent with the literature. Reports of lower education levels and higher comorbidities among LBCS compared to the NLW BCS echo findings in other studies (Yanez et al., 2011) . However, d espite lower education and higher comorbidities, the LBCS reported higher SWB and BF than their NLWBCS counterparts . These findings reflect the Latino Paradox, indicating that SWB and BF might serve LBCS in di stinct ways during survivorship, esp ecially long term survivorship . Findings additionally showed that SWB was an independent predictor of HRQOL and BF, respectively. However, the study also demonstrated null findings that contradicted study hypotheses, such that BF was not a significant pred ictor of HRQOL, and PTSS was a not a significant predictor of either HRQOL or BF in the LBCS sample. Overall, the study findings might have important implications for the improvement or development of oncology programs and interventions with LBCS, as well as future research with LBCS. With regard to SWB, researchers and clinicians should consider how LBCS might turn to spirituality as a coping strategy in ways different from NLWBCS, who have also been found to use spirituality as a coping strategy. In this way, high SWB might serve as a protective factor for LBCS and NLWBCS alike. S pirituality uniquely permeate s the roots of Latino cultures and provide s a rich texture to daily life ( Hunter Hern‡ndez et al., 2015; Gallo et al., 2009 ); thus, it is possible th at SWB serves a unique function for LBCS. Spirit uality and religiosity can facilitate an individual's intrinsic relationship with God or a higher power, as well as her extrinsic relationships with family and community. Spirituality and religiosity

PAGE 46

! ! ! 38 also hav e multidimensional relationships with other Latino cultural values, such that Latino individuals , in general, experience and express multiple values simultaneously (Campesino, 2009). Given all this , it is possible that SWB for LBCS significantly reinforces and is, in turn, reinforced by experiences that si multaneously promote other cultural values , such as personalismo (e.g., warmth, closeness, and empathy in relationships) and familismo (e.g., mutual commitment and loyalty among family members) . Another critical and relevant Latino cultural concept is colectivismo (i.e., collectivism) . Collectivism may be fostered through spiritual activities, such that LBCS might find that spiritual activities, especially those involving a spiritual community, can allow them to lean on their relationship s with the spiritual community and God. However, on an individual level, LBCS might experience peace and comfort through such spiritual coping activities as prayer. LBCS might additionally reflect upon the meaning of life and their purpose in life, both of which are concepts that individuals might ponder when facing their thoughts and feelings about the potential return of a terminal illness (Wildes, Miller, San Miguel de Majors & Ramirez, 2009). In this way, spirituality m ight serve as a channel through which LBCS can additionally reflect upon other aspects of their breast cancer experience from diagnosis to survivorship, including any losses or benefits they might have experienced. BF or perceived benefit, as previously described, can serve as a coping strategy for breast cancer survivors. Study findings showed that SWB was a significant, independent predictor of BF in the LBCS sample. LBCS might be more likely to perceive benefits from the breast cancer experience if the y consider how their relationships with God, family, friends, and community members might have been positively influenced. More specifically,

PAGE 47

! ! ! 39 LBCS who experience higher SWB might also lean more into their faith and community than those who experience lower SWB, and as a result, LBCS with higher SWB might derive more benefit from the breast cancer experience. Given the s tudy findings, researchers and clinicians might consider tailoring psychosocial interventions with LBCS to potential, demographically defi ned characteristics as lower educa tion and higher comorbidities. More specifically, researchers and clinicians can work to match the literacy level of materials to the literacy level of the group. Researchers and clinicians can also integrate information a bout and accommodate the tangible needs relevant to common comorbidities, such as providing healthy refreshments to diabetic patients or bariatric chairs for overweight patients. Additionally, researchers and clinicians can emphasize spirituality as a copi ng strategy in psychosocial interventions with LBCS. Oncology programs might further seek to establish connections with spiritual leaders and organizations within the hospital system or the community at large, in order to better expand their psychosocial s ervices beyond traditional psychotherapy or counseling interventions, especially for LBCS in long term survivorship who might be more focused on maintaining overall health and wellbeing than adjusting to survivorship. With regard to null study findings, p revious study findings in general BCS samples have found BF to be a significant predictor of HRQOL. However, the concept of BF is generally not well understood, let alone in LBCS. The literature suggests that BF is a positive experience that leads to posit ive outcomes (Campbell & Woodgate, 2015; Pascoe & Edvardsson, 2013) . However, much of this research involves NLW populations in the U.S. , it is possible that BF is a uniquely American concept that might be expressed and understood differently for LBCS comp ared to NLWBCS.

PAGE 48

! ! ! 40 R esearchers and clinicians should carefully consider how individuals who traverse challenging circumstances perceive Ôbenefit' as positive or negative . For example, LBCS who adhere strongly to Christian or Catholic perspective s , might vie w an adverse experience, such as b reast cancer as a punishment for having sinned. Some LBCS might perceive the breast cancer diagnosis as a divine invitation to repent and improve their lifestyle, interpersonal relationships, relationship to the community, as well as other areas of life. Individuals adhering to such beliefs might experience and view this type of benefit either positively or negatively. In this way, LBCS might derive benefit from the breast cancer experience through a distinctly spiritual or religious mechanism that is further influenced by the individual's particular cultural context. Given all this, future research is needed to unpack the concept of BF among diverse, ethno cultural population s, including LBCS . Q ualitative methods might be e specially helpful to better unpack the concept of BF among LBCS . With regard to PTSS and HRQOL , the results of the ANCOVAs did not support the hypotheses or previous research, in which NLW BCS tend to report higher perceived social support and HRQOL than their minority counterparts, including LBCS (Garcia Jimenez et al., 2014; Graves et al., 2012; Martinez Ramos et al., 2013; Yanez et al., 2011). However, bivariate correlations showed that race was correlated with HRQOL, which suggests that differences mi ght exist between the NLWBCS and LBCS groups. Thus, it is possible that the inclusion of covariates may have eliminated any statistically detectable differences. Alternatively, it may be possible that the lack of differences between the NLWBCS and LBCS on PTSS and HRQOL w e re due to higher homogeneity across the racial/ethnic groups in general , especially given that most of the participants were either married or living as married , regardless of race/ethnicity . It is possible that the experience of having a life partner

PAGE 49

! ! ! 41 might inherently involve either the direct experience of tangible social support or the perception that it is available if needed. Hirschman & Bourjolly (2005) conducted qualitative interviews with 33 women with breast cancer, in which the maj ority of women reported that their partner and mother were their main sources of tangible social support. Regardless, the findings on PTSS help add to the literature, because breast cancer survivorship research, in general, has not yet focused on the influ ence of PTSS on HRQOL or BF. Future research should target a more diverse sample of participants with regard to marital status, in order to better observe any differences between the LBCS and NLWBCS on PTSS, and whether or not PTSS is associated with HRQOL or BF in single (i.e., never married, widowed, divorced) LBCS and NLWBCS . With regard to HRQOL in the LBCS , results showed that marital status and comorbidities were significant predictors of HRQOL. Future research and interventions aiming to improve H RQOL in LBCS might consider whether or not marital status and comorbidities are unique predictor s of HRQOL for LBCS compared to other racial/ethnic BCS. Unique predictors of HRQOL might also be useful for researchers and clinicians developing interventions that aim to improve HRQOL in LBCS. It is possible that marital status is in important predictor of HRQOL for LBCS due to the Latino cultural value s of familismo (focus on family) and family/social support , because family/social support often include s part ners (A–ez et al., 2005; Gallo et al., 2009) . In general, familismo and family/social support have been found to be protective, such that they often help to reduce risk and improve resilience and health outcomes in Latino populations (Hunter Hern‡ndez et a l., 2015; Gallo et al., 2009) . Therefore, marital or Ôpartner' status might play a key role in HRQOL for LBCS .

PAGE 50

! ! ! 42 Overall, this study had several strengths and potential limitations. With regard to strengths, SUNSHINE was a population based epidemiological study, such that the state wide cancer registries have the opportunity to capture data from virtually all persons diagnosed with and treated for cancer. State wide cancer registries can also contain diverse data across this population. In this way, researc hers were able to identify potential participants from a participant pool that represents the entire population of breast cancer survivors across the particular states of Colorado and Arizona, rather than, for example, specific hospital sites. Given that t his study utilized such population based data, study results are more likely to be generalizable to the LBCS population of Colorado and Arizona. Researchers were also able to observe previously collected data, such as demographic and medical data, across s everal years. Another key advantage of this study was that researchers were able to focus on long term breast cancer survivorship. Participants completed the SUNSHINE surveys at least 6 years after their last treatment (Sedjo et al., 2013). Research has te nded to focus on short term survivorship, especially with regard to HRQOL. Therefore, the study results might particularly help improve understanding of the factors that contribute to HRQOL and BF among LBCS in long term survivorship specifically. However, study results should be interpreted caution due to potential limitations . A potential threat to internal validity is the possibility that some participants were influenced by the participant response bias of social desirability . Some Latina participants m ay have been hesitant to express negative emotions for fear of being stigmatized, discriminated socially, embarrassed, or labeled according to a mental illness (e.g., "depressed") (Vega, 2010; Nadeem et al., 2007). Therefore, some Latina participants might have reported higher SWB and higher HRQOL on the respective measures. Latina participants might have also

PAGE 51

! ! ! 43 attempted to respond favorably if they completed study interviews over the phone with the research assistants, or if a relative or friend helped them complete an interview(s). Participants might have wanted the relative or friend to view them in a positive light, or the relative or friend might have encouraged participants to respond more optimistically. This study was unable to assess how mental heal th stigma may have influenced self report measures within the interviews, because none of the measures involve questions on participants' perspectives towards mental illness, such as depression or anxiety. In future studies, it would be worth taking accoun t of mental health stigma within Latino communities . This could potentially be done by asking participants about their perspectives and beliefs related to mental illness, especially depression and anxiety. Such questions may be more feasibly posed through qualitative versus quantitative interviews, because participants could be invited to elaborate upon their perspectives. In contrast, quantitative measures would need to be improved. It is possible that responses to questions regarding mental health stigma within Latino communities may not be easily quantified , such as with the use of Likert style responses . M ental health stigma within Latino communities might also be better conceptualized as a multidimensional construct. Although it is hoped that the resul ts will generalize to LBCS across the U.S. , there may be limitations to external validity, particularly with regard to the structural component. In terms of the population sample, the majority of Latinos living in Colorado and Arizona are Mexican or Mexica n American, which suggests that the results may not necessarily generalize to other Latino subgroups. It is also important to consider regional differences across the U.S . Because the LBCS sample is from the Southwest U.S. (i.e., Colorado and Arizona), the y might exhibit differences in , for example, lifestyle, subculture, healthcare

PAGE 52

! ! ! 44 access, access to food, ethic/racial diversity . As a result, the LBCS sample may not generalize to LBCS living in other regions of the United States (e.g., metropolitan Chicago, metropolitan New York City, metropolitan Los Angeles) . Additionally, the state wide cancer registries might have a limited number of underserved and undocumented LBCS (Risendal et al., 2015), which could include those who lack or have limited access to qu ality healthcare, or have lower education. A potential lack of underserved, undocumented LBCS might explain the restriction in range on education levels among the LBCS. Latinos in general in the U.S. are significantly less likely to graduate from high scho ol, let alone pursue higher education, compared to their NLW counterparts ( Krogstad, 2016 ). However, the restriction in range on such demographic variables as education and marital status might be due to selection bias. Although the sample was a population based sample and was derived from state wide cancer registries, it is possible that the LBCS who agreed to participate tended to be those of, for example, higher education and either married or living as married. Despite the limitations , however, the st udy findings not only support previous research on racial/ethnic disparities in breast cancer survivorship, but they might also contribute to improving our understanding of potential predictors of HRQOL and BF in LBCS. With improved understanding, research ers and clinicians might soon begin to fill the need for effective post treatment interventions for LBCS throughout the Southwest United States and beyond .

PAGE 53

! ! ! 45 REFERENCES A–ez, L. M., Paris, J., Manuel, Bedregal, L. E., Davidson, L., & Grilo, C. M. (2005). Application of cultural constructs in the care of first generation Latino clients in a community mental health setting. Journal of Psychiatric Practice,11 (4), 221 230. doi:10.1097/00131746 200507000 00002 Antoni, M. H., Lehman, J. M., Kilbourn, K. M., Bo yers, A. E., Culver, J. L., Alferi, S. M.. . Carver, C. S. (2001). Cognitive behavioral stress management intervention decreases the prevalence of depression and enhances benefit finding among women under treatment for early stage breast cancer. Health Psy chology, 20 (1), 20 32. doi:10.1037/0278 6133.20.1.20 Ashing, K., Rosales, M., Lai, L., & Hurria, A. (2014). Occurrence of comorbidities among African American and Latina breast cancer survivors. Journal of Cancer Survivorship , 8 (2), 312 318. doi:10.1007/s 11764 014 0342 x Ashing Giwa, K. T., Kim, J., & Tejero, J. S. (2008). Measuring quality of life among cervical cancer survivors: Preliminary assessment of instrumentation validity in a cross cultural study. Quality of Life Research, 17 (1), 147 157. doi:10 .1007/s11136 007 9276 3 Ashing Giwa, K., J. Tejero, et al. (2007). Examining predictive Models of HRQOL in a population based, multi ethnic sample of women with breast carcinoma. Quality of Life Research, 16 , 413 428. Askim Lovseth, M.K. & Aldana, A. (2 010). Looking beyond "affordable" health care: Cultural understanding and sensitivity Ñ necessities in addressing the health care disparities of the U.S. Hispanic population. Health Marketing Quarterly, 27:4, 354 387, DOI: 10.1080/07359683.2010.519990 Brady , M. J., Peterman, A. H., Fitchett, G., Mo, M., & Cella, D. (1999). A case for including spirituality in quality of life measurement in oncology. Psycho ! oncology, 8 (5), 417 428. doi:10.1002/(SICI)1099 1611(199909/10)8:5<417::AID PON398>3.0.CO;2 4 Brady, M . J., Cella, D. F., Mo, F., Bonomi, A. E., Tulsky, D. S., Lloyd, S. R.. . Shiomoto, G. (1997). Reliability and validity of the functional assessment of cancer therapy breast quality of life instrument. Journal of Clinical Oncology, 15 (3), 974 986. Bravema n, P. A., Cubbin, C., Egerter, S., Williams, D. R., & Pamuk, E. (2010). Socioeconomic disparities in health in the United States: What the patterns tell us. American Journal of Public Health, 100 (S1), S186 S196. doi:10.2105/AJPH.2009.166082

PAGE 54

! ! ! 46 Bredle, J. M., Salsman, J. M., Debb, S. M., Arnold, B. J. & Cella, D. (2011). Spiritual wellbeing as a component of health related quality of life: the functional assessment of chronic illness therapy Ñ spiritual wellbeing scale (FACIT Sp). Religions , 2 (77), 77 94. doi:10 .3390/rel2010077 Bronfenbrenner, U. (1994). Ecological models of human development. In International Encyclopedia of Education , Vol. 2, 2 nd edition. Oxford: Elsevier. Brookings J. B., Bolton B. (1988). Confirmatory factor analysis of the Interpersonal Support Evaluation List. American Journal of Community Psychology , 16 , 137 147 Campbell Enns, H. & Woodgate, R. (2015). The psychosocial experiences of women with breast cancer across the lifespan: a systematic review protocol. The Joanna Briggs Institute Database of Systematic Reviews Implementation Reports . 13 (1):112 21. doi: 10.11124/jbisrir 2015 1795. Campesino, M. (2009). Spirituality Among Latinas/os Implications of Culture in Conceptualization and Measurement. Advances in Nursing Science. 29 (1): 69 81. Carver, C. S. (2013). The Benefit Finding Scale for breast cancer. Measurement Instrument Database for the Social Science. Retrieved from www.midss.ie Cella, D. F., Tulsky, D. S., Gray, G., Sarafian, B., Linn, E., Bonomi, A... & Brannon, J. (1993). The functional assessment of cancer therapy scale: Development and validation of the general measure. Journal of Clinical Oncology,11 (3), 570. Cella, D., Hernandez, L., Bonomi, A. E., Corona, M., Vaquero, M., Shiomoto, G., & Baez, L. (1998). Spanish language translation and initial validation of the functional assessment of cancer therapy quality of life instrument. Medical Care, 36 (9), 1407 1418. doi:10.1097/00005650 199809000 00012 Chopra, I. & Kamal, K.M. (2012). A s ystematic review of quality of life instruments in long term breast cancer survivors. Health and Quality of Life Outcomes , 10 (14). doi: 10.1186/1477 7525 10 14 Cohen, S., & Hoberman, H. (1983). Positive events and social supports as buffers of life change stress. Journal of Applied Social Psychology , 13 , 99 125. Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. M. (1985). Measuring the functional components of social support. In I. G. Sarason & B. R. Sarason (Eds.), Social support: Theory, research, and applications (pp. 73 94). The Hague, The Netherlands: Martinus Nijhoff.

PAGE 55

! ! ! 47 Crookes, D. M., Shelton, R. C., Tehranifar, P., Aycinena, C., Gaffney, A. O., Koch, P.. . Greenlee, H. (2016). Social networks and social support for healthy eating among Latina breast cancer survivors: Implications for social and behavioral interventions. Journal of Cancer Survivorship, 10 (2), 291 301. doi:10.1007/s11764 015 0475 6 Davis, L. E., Engel, R. J., & SpringerLink (Online service). (2011). "Chapter 2: Hispanics." Measu ring race and ethnicity . New York: Springer. doi: 10.1007/978 1 4419 6697 1_2 Gallo, L. C., Penedo, F. J., Espinosa de los Monteros, Karla, & Arguelles, W. (2009). Resiliency in the face of disadvantage: Do Hispanic cultural characteristics protect healt h outcomes? Journal of Personality, 77 (6), 1707 1746. doi:10.1111/j.1467 6494.2009.00598.x Garc’a Jimenez, M., Santoyo Olsson, J., Ortiz, C., Lahiff, M., Sokal Gutierrez, K., & N‡poles, A. M. (2014). Acculturation, inner peace, cancer self efficacy, and s elf rated health among Latina breast cancer survivors. Journal of Health Care for the Poor and Underserved, 25 (4), 1586 1602. doi:10.1353/hpu.2014.0158 Garc’a Jimenez, M. (2013). Culture and Self rated Health Among Latina Breast Cancer Survivors: The Role of Cancer Self efficacy and Spiritual Wellbeing (Master's thesis). Retrieved from eScholarship University of California Graves, K. D., Jensen, R. E., Ca–ar, J., Perret Gentil, M., Leventhal, K., Gonzalez, F... Mandelblatt, J. (2012). Through the lens of culture: Quality of life among Latina breast cancer survivors. Breast Cancer Research and Treatment , 136(2), 603 613. doi:10.1007/s10549 012 2291 2 Hegelson, V. S., Reynolds, K. A., & Tomich, P. L. (2006). A meta analytic review of benefit finding and gro wth. Journal of Consulting and Clinical Psychology, 74 (5), 797 816. doi:10.1037/0022 006X.74.5.797 Hirschman, K. B. & Bourjolly, J. N. (2005). How Do Tangible Supports Impact the Breast Cancer Experience? Social Work in Health Care, 41:1, 17 32, DOI: 10 .1300/J010v41n01_02 Hoddinott, S. N., & Bass, M. J. (1986). The Dillman total design survey method. Canadian Family Physician MŽdecin De Famille Canadien , 32 , 2366 2368. Hodge, D. R., Wolosin, R. J., & Bonifas, R. P. (2013). Addressing older Latinos' sp iritual needs in hospital settings: Identifying predictors of satisfaction. Advances in Social Work, 14 (1), 65 81. Hunter Hern‡ndez, M., Costas Mu–’z, R., & Gany, F. (2015). Missed opportunity: Spirituality as a bridge to resilience in Latinos with cancer . Journal of Religion and Health, 54 (6), 2367 2375. doi:10.1007/s10943 015 0020 y

PAGE 56

! ! ! 48 IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp. Janz, N.K., Friese, C.R., Li, Y., Graff, J.J., Hamilton, A.S., & Hawley, S.T. (2014). Emotional wellbeing years post treatment for breast cancer: prospective, multi ethnic, and population based analysis. Journal of Cancer Survivorship , 8 , 131 Ð 142. doi: 10.1007/s11764 013 0309 3 Jansen, L., Hoffmeister, M., Chang Claude, J., Brenner , H. & Arndt, V. (2011). Benefit finding and post traumatic growth in long term colorectal cancer survivors: prevalence, determinants, and associations with quality of life. British Journal of Cancer , 105 , 1158 Ð 1165. doi:10.1038/bjc.2011.335 Kagawa Sing er, M., Padilla, G. V., & Ashing Giwa, K. (2010). Health related quality of life and culture. Seminars in Oncology Nursing,26 (1), 59 67. doi:10.1016/j.soncn.2009.11.008 Kroenke, C. H., Quesenberry, C., Kwan, M. L., Sweeney, C., Castillo, A., & Caan, B. J. (2013a). Social networks, social support, and burden in relationships, and mortality after breast cancer diagnosis in the life after breast cancer epidemiology (LACE) study. Breast Cancer Research and Treatment, 137 (1), 261 271. doi:10.1007/s10549 012 225 3 8 Kroenke, C. H., Kwan, M. L., Neugut, A. I., Ergas, I. J., Wright, J. D., Caan, B. J.. . Kushi, L. H. (2013b). Social networks, social support mechanisms, and quality of life after breast cancer diagnosis. Breast Cancer Research and Treatment, 139 (2), 515 527. doi:10.1007/s 105 49 013 2477 2 Kroenke C. H., Michael Y., Tindle H., Gage E., Chlebowski R., Garcia L., Messina C., Manson J. E., Caan B. J. (2012) Social networks, social support and burden in relationships, and mortality after breast cancer diag nosis. Breast Cancer Res Treat 133 (1): 375 Ð 385 Krogstad, J.M. (2016). 5 Facts About Latinos and Education. The Pew . Retrieved from: http://www.pewresearch. org/fact tank/2016/07/28/5 facts about latinos and education/ Levine, E. G., Yoo, G., Aviv, C., Ewing, C., & Au, A. (2007). Ethnicity and spirituality in breast cancer survivors. Journal of Cancer Survivorship, 1 (3), 212 225. doi:10.1007/s11764 007 0024 z L—pez Class, M., Perret Gentil, M., Kreling, B., Caicedo, L., Mandelblatt, J., & Graves, K. D. (2011). Quality of life among immigrant Latina breast cancer survivors: Realities of culture and enhancing cancer care. Journal of Cancer Education , 26(4), 72 4 733. doi:10.1007/s13187 011 0249 4

PAGE 57

! ! ! 49 Maly, R. C., Stein, J. A., Umezawa, Y., Leake, B., & Anglin, M. D. (2008). Racial/Ethnic differences in breast cancer outcomes among older patients: Effects of physician communication and patient empowerment. Health Ps ychology, 27 (6), 728 736. doi:10.1037/0278 6133.27.6.728 Martinez Ramos, G. P., Biggs, M. J. G., & Lozano, Y. (2013). Quality of life of Latina breast cancer survivors: From silence to empowerment. Advances in Social Work, 14 (1), 82 101. Mickley, J., & S oeken, K. (1993). Religiousness and hope in Hispanic and Anglo American women with breast cancer. Oncology Nursing Forum, 20 (8), 1171. Nadeem, E., Lange, J. M., Edge, D., Fongwa, M., Belin, T., & Miranda, J. (2007). Does stigma keep poor young immigrant and U.S. born black and Latina women from seeking mental health care? Psychiatric Services (Washington, D.C.), 58 (12), 1547 1554. N‡poles, A. M., Ort’z, C., O'Brien, H., Sereno, A. B., & Kaplan, C. P. (2011). Coping resources and self rated health amo ng Latina breast cancer survivors. Oncology Nursing Forum , 38(5), 523. doi:10.1188/11.ONF.523 531 Ortega, A. N., Rodriguez, H. P., & Vargas Bustamante, A. (2015). Policy dilemmas in L atino health care and implementation of the affordable care act. Annual Review of Public Health, 36 (1), 525 544. doi:10.1146/annurev publhealth 031914 122421 Pew Research Center (2014). The shifting religious identify of Latinos in the United States: Nearly one in four Latinos are former Catholics. Pew Research Center . Retri eve from http://www.pewforum.org/2014/05/07/the shifting religious identity of latinos in the united states/ Pinheiro, P. S., Williams, M. , Miller, E. A., Easterday, S., Moonie, S., & Trapido, E. J. (2011). Cancer survival among Latinos and the Hispanic Paradox. Cancer Causes & Control, 22 (4), 553 561. doi:10.1007/s10552 011 9727 6 Pascoe, L., Edvardsson, D., UmeŒ universitet, Institutionen fšr omvŒrdnad, & Medicinska fakulteten. (2013). Benefit finding in cancer: A review of influencing factors and health outcomes. European Journal of Oncology Nursing: The Official Journal of European Oncology Nursing Society, 17 (6), 760. doi:10.1016/j.ejon .2013.03.005 Patel, M. I., Schupp, C. W., Gomez, S. L., Chang, E. T. & Wakelee, H. A. (2013). How do social factors explain outcomes in non Ð small cell lung cancer among Hispanics in California? Explaining the Hispanic Paradox. Journal of Clinical Oncology . 31 (28). doi: 10.1200/JCO.2012.48.6217 Rasheed, M. N., Rasheed, J. & Marley, J. A. (2011). Cultural values in relation to family structure. Family Therapy: Models and Techniques . (pp. 111 115). Thousand Oaks, CA: Sage Publications, Inc.

PAGE 58

! ! ! 50 Risendal, B. C ., Sedjo, R. L., Giuliano, A. R., Vadaparampil, S., Jacobsen, P. B., Kilbourn, K.. . Byers, T. (2015). Surveillance and beliefs about follow up care among long term breast cancer survivors: A comparison of primary care and oncology providers. Journal of Ca ncer Survivorship, 10 (1), 96 102. doi:10.1007/s11764 015 0454 y Ruiz, J. M., Campos, B. & Garcia, J. J. (2016). Special issue on Latino physical health: Disparities, paradoxes, and future directions. (2016). Journal of Latina/o Psychology, 4 (2), 61 66. do i:10.1037/lat0000065 Schwartz, S. J., Unger, J. B., Zamboanga, B. L., & Szapocznik, J. (2010). Rethinking the concept of acculturation: Implications for theory and research. American Psychologist, 65 (4), 237 251. doi:10.1037/a0019330 Sedjo RL, Hines LM , Byers T, Giuliano AR, Marcus A, Vadaparampil S, Jacobsen P, Kilbourn K, Feser WJ, Risendal BC (2013). Long term weight gain among Hispanic and non Hispanic White women with and without breast cancer. Nutrition and Cancer . 65: 34 42. PMID 23368911 DOI: 10.1080/01635581.2013.741750 Siegel, R. L., Fedewa, S. A., Miller, K. D., Goding ! Sauer, A., Pinhei ro, P. S., Martinez ! Tyson, D., & Jemal, A. (2015). Cancer statistics for Hispanics/Latinos, 2015. CA: A Cancer Journal for Clinicians, 65 (6), 457 480. doi:10.3322/caac.21314 Tammemagi, C.M., Nerenz, D., Neslund Dudas, C., Feldkamp, C. & Nathanson, D. (200 5). Comorbidity and survival disparities among black and white patients with breast cancer. The Journal of the American Medical Association. 294 (14). Tomich, P. L., & Helgeson, V. S. (2004). Is finding something good in the bad always good? benefit findin g among women with breast cancer. Health Psychology, 23 (1), 16 23. doi:10.1037/0278 6133.23.1.16 Uchino, B. N. (2004). Social support and physical health: Understanding the health consequences of relationships . New Haven: Yale University Press. Universit y of Manitoba (2016). Concept: Charlson Comorbidity Index. Retrieved from: http://mchp appserv.cpe.umanitoba.ca/viewConcept.php?conceptID=1098 U.S. Census Bureau (2015). Hispanic Heritage Month 2015. Profile American Facts for Features . Retrieved from: https://www.census.gov/content/dam/Census/new sroom/facts for features/2015/cb15 ff18_hispanic_heritage_month.pdf Vega, W. (2010). Addressing stigma of depression in Latino primary care patients. General Hospital Psychiatry, 32 (2), 182 191. doi:10.1016/j.genhosppsych.2009.10.008

PAGE 59

! ! ! 51 Viruell Fuentes, E. A., & Schulz, A. J. (2009). Toward a dynamic conceptualization of social ties and context: Implications for understanding immigrant and Latino health. American Journal of Public Health, 99 (12), 2167 2175. doi:10.2105/AJPH.2008.158956 Weaver, K. E., Lla bre, M. M., Lechner, S. C., Penedo, F., & Antoni, M. H. (2008). Comparing unidimensional and multidimensional models of benefit finding in breast and prostate cancer. Quality of Life Research, 17 (5), 771 781. doi:10.1007/s11136 008 9348 z Wheeler, S.B., R eeder Hayes, K.E. & Carey, L.A. (2013). Disparities in breast cancer treatment and outcomes: Biological, social, and health system determinants and opportunities for research. The Oncologist , 18 , 986 Ð 993. doi: http://dx.doi.org/10.1634/theoncologist.2013 0243 Wildes, K. A., Miller, A. R., San Miguel de Majors, S. & Ramirez, A. G. (2009). The religiosity/spirituality of Latina breast cancer survivors and influence on health related quality of life. Psychooncology . 18 (8): 831 Ð 840. doi: 10.1002/pon.1475 Wu, A. H., Kurian, A. W., Kwan, M. L., John, E. M., Lu, Y., Keegan, T. H. M.. . Vigen, C. (2015). Diabetes and other comorbidities in breast cancer survival by race/ethnicity: The California br east cancer survivorship consortium (CBCSC). Cancer Epidemiology, Biomarkers & Prevention: A Publication of the American Association for Cancer Research, Cosponsored by the American Society of Preventive Oncology , 24 (2), 361 368. doi:10.1158/1055 9965.EPI 14 1140 Yanez, B., Thompson, E. H., & Stanton, A. L. (2011). Quality of life among Latina breast cancer patients: A systematic review of the literature. Journal of Cancer Survivorship, 5 (2), 191 207. doi:10.1007/s11764 011 0171 0