University of Southern California Dissertations and Theses

Adaptation to a life -threatening diagnosis: Dispositional optimism and pessimism and posttraumatic growth among patients with HIV

(USC Thesis Other)

Adaptation to a life -threatening diagnosis: Dispositional optimism and pessimism and posttraumatic growth among patients with HIV

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UM I a complete manuscript and there are missing pages, these w ill be noted. Also, if unauthorized copyright material had to be removed, a note w ill indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ADAPTATION TO A LIFE THREATENING DIAGNOSIS: DISPOSITIONAL OPTIMISM AND PESSIMISM AND POSTTRAUMATIC GROWTH AMONG PATIENTS WITH HIV by Joel Ellis Milam A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (PREVENTIVE MEDICINE) May 2002 Copyright 2002 Joel Ellis Milam Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3073815 ____ < g ) UMI UMI Microform 3073815 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA The Graduate School University Park LOS ANGELES, CALIFORNIA 90089-1695 This dissertation, w ritten b y Under the direction o f Dissertation Committee, and approved b y all its members, has been presented to and accepted b y The Graduate School, in partial fulfillm ent o f requirements for the degree o f DOCTOR OF PHILOSOPHY Dean o f Graduate Studies D ate May 10, 2002 DISSER TA H O N COMMITTEE Chairperson ' S C 7 **i ^ * * 1 / ' v 7 /< f Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS I wish to thank the following people for their invaluable support: Dr. Jean Richardson, my mentor, who has continuously believed in my abilities and gave me the opportunity to pursue this line of research; Dr. Gary Marks, with whom I originally interviewed with for admission to USC, who convinced me that USC was the place for me, and who has provided me with continuous encouragement since; Sue Stoyanoff, without whom this study would not have been completed; Mamy Barovich, who consistently helped me navigate graduate school bureaucracy and always had an answer to my questions; My committee members, role models whom provided inspiration through their challenging questions and suggestions; All o f the clinics, staff, and study participants who embraced the larger “Partnership for Health” study; My wife, Shelly, who is always there for me and whose confidence, stability, and love continuously inspires me; My family, and friends who—although they did not always understand everything I was working on—enthusiastically supported me. This research was supported by NIMH RO1MH57208; The author was supported by NCI T32CA09492. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS Acknowledgements ii List of Tables iv List of Figures v Abstract vii Introduction 1 Background and Significance 3 Posttraumatic Growth 3 Dispositional Optimism and Pessimism 13 Positive Beliefs and HIV 3 1 Present Study 34 Method 41 Results 50 Discussion 100 References 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES 1 . Characteristics o f the Sample at Baseline (n=83 5) 53 2. Levels of Posttraumatic Growth Endorsement at Baseline (n=83 5) 55 3. Mean (SD) Differences in PTG, Optimism, and Pessimism at baseline (n=835) 57 4. Correlations Between Major Variables (n=83 5) 59 5. Multiple Regression Analyses o f Disease Status 64 6. Multiple Regression Analyses of Depression 79 7. Multiple Regression Analyses of PTG 86 8. Multiple Regression Analyses o f Optimism and Pessimism 90 9. Multiple Regression Analyses of ART Adherence and Side Effect Reporting 93 10. Summary o f Multiple Regression Analyses 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES 1. Samples 51 2. PTG was associated with lower levels o f viral load at baseline among those with low CD4 counts 65 3. PTG had a positive impact on viral load at follow-up among those scoring low in pessimism 65 4. Optimism had a positive impact on viral load at follow-up for females 66 5. Females scoring high in pessimism had the highest CD4 counts at baseline 68 6. Those high in depression and pessimism had the lowest CD4 counts at baseline 68 7. Hispanics who experienced PTG had a significant increase in CD4 counts over time 69 8. PTG had a beneficial influence on CD4 count overtime among those with moderate levels of optimism 70 9. Those with undetectable viral loads and low pessimism scores had the highest CD4 counts over time 70 10. Relationship between changes in PTG and CD4 counts over time among Hispanics 72 11. Relationship between changes in PTG and CD4 counts over time 72 12. Females scoring low in optimism had the highest levels of depression at baseline 80 13. Those with low optimism scores and detectable viral loads had the highest depression levels at baseline 80 14. Those scoring low in depression and pessimism at time 1 had the lowest depression scores at time 2 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15. Among those scoring high in optimism at baseline, those who started ART use were more likely to become depressed 16. Relationship between PTG and depression over time 17. Those scoring high in pessimism with detectable viral loads had the lowest levels of adherence 18. Those scoring high in pessimism and depression had the lowest levels of adherence 19. Pessimism had an inverse relationship with adherence for all ethnicities except Hispanic 20. There was a positive relationship between optimism and adherence for females 21. The relationship between optimism and side effect reporting was greatest for African-Americans Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Although a life-threatening diagnosis can lead to many negative outcomes, there are also positive outcomes and protective factors that buffer against mental and physical illness. Posttraumatic growth (PTG; the construing o f benefits from a negative life event) and dispositional optimism/pessimism are constructs that play important roles during psychological and physical challenge. However, the extent to which these strengths aid in the adaptation to disease is unclear. Furthermore, the processes underlying these constructs need further clarification. Among a diverse sample of HIV patients (n=835 at baseline) from 6 outpatient clinics located throughout California, this longitudinal study examined the relationships between PTG, dispositional optimism/pessimism and; disease status/progression, depressive symptoms, medication adherence, and side effect reporting. Potential mediators and moderators of these relationships were explored. Disease progression was indicated by higher levels of viral load and/or lower levels of CD4 counts over time. Hypotheses were examined with hierarchical multiple regression analyses that included important covariates (e.g., age, antiretroviral use, baseline levels o f CD4 count). The major hypotheses were partially supported and indicated that: a) PTG is prevalent; 59% experienced moderate to high levels o f PTG, b) PTG had a negative association with viral load, a positive association with CD4 counts among Hispanics and those with moderate optimism scores, and a negative association with the development of depressive symptoms, c) Pessimism was positively associated with disease progression, the development of depressive symptoms, and lower levels of vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. adherence, and d) Optimism had a negative association with viral load for females (a relationship mediated by higher levels of medication adherence) and a positive association with depressive symptoms among those who started antiretroviral therapy, suggesting a negative ramification of holding positive expectancies that are disconfirmed. Health behaviors did not explain the majority of these relationships. These data indicate that, although their underlying processes are complex, PTG and optimism/pessimism operate independently and aid in the adaptation and adjustment to HIV/AIDS. Research and clinical health interventions should include efforts to foster PTG and optimism and to inhibit pessimism. V lll Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION The past decade has been witness to substantial progress in addressing the dynamic relationship between physical health and a multitude o f psychosocial factors. Partially explaining this advancement is the ease with which theories that were originally developed to account for mental health have been used to promote and understand physical health (Adler & Matthews, 1994). Researchers and physicians alike have acknowledged the importance o f examining patients as a whole rather than as the sum o f separate systems, emphasizing the interdependence between physical and mental health. HIV is a disease where there is much need for this type o f integration. That is, due to antiretroviral therapies (ART), people with HIV or AIDS who have dealt with the prospect o f impending death, now have the possibility to live a much longer life. Thus, living with HIV/AIDS as a chronic illness can have a profound impact on one’s priorities and outlook on life. These changes in life priorities can impact how one adjusts to their illness, from infection to disease progression/stability and death. Therefore, the dynamic relationship between a person’s psychological and physical status in the context o f HIV adaptation warrants examination. One important mental health construct that has received much attention in relation to health is dispositional optimism/pessimism. Likewise, a second important construct that has been documented, but has not been as rigorously examined in relation to health, is posttraumatic growth (PTG); the construing o f benefits from a traumatic event, such as a life-threatening diagnosis. Both o f these constructs play 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. important roles in the adaptation to disease. The purpose o f this dissertation study is to determine levels o f PTG and dispositional optimism among a sample of people living with HIV and AIDS and to examine the relationships between these constructs and mental, physical, and behavioral health indicators. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BACKGROUND AND SIGNIFICANCE Posttraumatic Growth Existentialism indicates that suffering allows for the contemplation of life’s most important questions, most notably whether life has any meaning. Paradoxically, those who accept these existential challenges are likely to experience both dread/anguish and meaningful personal growth. A traumatic event, such as a diagnosis with a life threatening illness, is such a catalyst where one can become acutely aware of these questions and personally consider their implications. Thus, in addition to simply going to the doctor and receiving treatment, the healing process can include forms of individual transformation, such as behavioral changes and psychological shifts in perception that change one’s sense of identity and priorities in life. To date, the bulk of the evidence for this type of posttraumatic growth has been anecdotal, ranging from popular media accounts to personal testimonies. Furthermore, the majority of the scientific models regarding adaptation to stressful or traumatic life events do not easily accommodate for the positive changes that have been observed in response to traumatic stress and illness. This is because PTG is considered more than just adaptation or being resilient, it is thriving; the moving beyond one’s original levels of functioning (O’leary & Ickovicks, 1995). An existential framework can successfully explain how PTG goes beyond simple adjustment or adaptation. That is, through the process o f dealing with existential issues, individuals can actually function at higher levels after a trauma compared to 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the levels where they were before the trauma (e.g., Farran, 1997; Yalom & Lieberman, 1991). Definition of PTG If you were diagnosed with a life threatening illness, how would you change the way you’re presently living your life? It is the positive answers to this question, the things that make life more personally meaningful, that represents PTG. Thus, PTG is the presence of positive changes following a traumatic event. These include both intrapersonal (e.g., an increased appreciation of life, changes in life priorities, spirituality, etc.) and interpersonal (e.g., improvements in relationships with family, increased expressiveness) changes. Intrapersonal changes Intrapersonal changes include increases in self-reliance, personal strength, and appreciation of one’s vulnerability. Other changes include the development of greater patience, tolerance, empathy, altruism, and courage (Affleck & Tennen, 1996). Often times, a person becomes stronger through their belief that “if I can get through this, I can get through anything.” For example, in a sample o f cancer patients, the most common reported positive change was feeling stronger and more self-assured (Collins, Taylor, & Skokan, 1990). Among survivors o f a sinking ship, 83% reported that they felt more experienced about life (Joseph, Williams, & Yule, 1993). In a sample o f gay men, HIV became a developmental trigger to discover internal resources that had never been put to the test (Schwartzberg, 1994). 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appreciation of one’s vulnerability may appear to be a paradox in the area of PTG. That is, it is difficult to conceive any positive aspects o f viewing oneself as vulnerable. However, vulnerability plays an essential role in the process of PTG (which is discussed in the next section). Life threatening illnesses raise levels of existential awareness, which allows for a greater appreciation of life. For example, in a sample of cancer patients, 45% thought differently about the future as a result of their cancer, indicating “that they valued life more and had a greater awareness of their mortality” (Charles, Sellick, Montesanto, & Mohide, 1996, p. 68). Thus, a changed philosophy of life is an aspect of PTG. In these cases, schemas are revised, priorities are rearranged, and new world views are adopted. Oftentimes, a ‘here and now’ focus is embraced. For example, of women who reported changes since cancer diagnoses, 60% reported changing their priorities—such as enjoying life more (Taylor, Lichtman, & Wood, 1984). Among ship sinking survivors, 94% reported that they no longer took life for granted and 71% reported to be living each day to the fullest (Joseph et al., 1993). A person’s spiritual life can also change. Spirituality/religiosity has been shown to both strengthen and weaken after trauma (Tedeschi & Calhoun, 1995), although this may depend on a one’s pre-trauma level o f spirituality. Interpersonal changes Interpersonal changes include increases in self-disclosure, emotional expressiveness, compassion, empathy, and efforts in relationships. Trauma can allow one to realize the importance of honest communication and genuine 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. relationships. For example, in a sample of cancer patients, 19% indicated that they had experienced improved relationships as a result of their cancer (Charles et al., 1996). Likewise, greater family support (more cohesion and expressiveness, less conflict) has been found among parents after a child’s death (Birenbaum & Robinson, 1991). Similarly, after cancer treatment, couples report that their relationships were strengthened through the experience (Gritz, et al., 1990). That is, there was increased intimacy, communication, and sensitivity to feelings. Perhaps the most striking example is that of cancer patients who are thankful they have cancer as opposed to other, more immediate diseases (e.g., heart attack, stroke, etc.) because it gives time to appreciate the love and tenderness that is possible between people (Mumford, 1997). Trauma can also spur community action. For example, HIV can be a motivating source for a sense of membership or belonging to a wider community (Schwartzberg, 1994). Thus, the interpersonal aspects of PTG can have a wide range of influence. For the purposes o f this research, PTG is defined as positive changes following HIV diagnosis in the following areas: appreciation of life, life priorities, spirituality, relationships, and self-reliance. Process of PTG It is evident that stress and trauma can provide a catalyst for growth, however the process through which this occurs is less clear. It does appear that the initial reduction o f invulnerability, or increase of vulnerability, after trauma plays an important role in subsequent growth. This shattering of assumptions (Janoff- 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bulman, 1992), or reduction of positive illusions (Taylor & Brown, 1988), appears to provide a window of realism in which one can concentrate on various existential matters (i.e., what is most important in life). Research reporting positive associations between PTG and the severity of the trauma (Aldwin, Levenson, & Spiro, 1994; Park, Cohen, & Murch, 1996; Tedeschi & Calhoun, 1996) and levels of existential awareness (Yalom & Lieberman, 1991) support this view. Thus, the strength of the trauma (i.e., ability to disrupt normal cognitive functioning) and heightened existential awareness allow for the cognitive processing necessary for PTG to occur. Time appears to play two important roles in the underlying process of PTG. First, preliminary evidence indicates that the chronological time elapsed since the occurrence of the trauma can be important in fostering PTG (Cordova, Cunningham, Carlson, & Andrykowski, 2001; Park et al., 1996), although others do not find this association (Tedeschi & Calhoun, 1996). While PTG can occur immediately after trauma, time may allow for proper rumination o f the important issues inherent to PTG. As one resumes their forward life trajectory, a new assumptive world is in put in place. This new assumptive world may be different, however it can theoretically be better than the old one. Second, the perception o f time plays a role in the cognitive processing underlying PTG. The notion that people adjust their goals according to the time they have left in life is the basis of socioemotional selectivity theory (see Carstensen, Isaacowitz, & Charles, 1999, for a review). This theory suggests that when time is viewed as open-ended, there is a premium on acquiring 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. information. Whereas when time is limited, short-term goals (notably emotional) take priority. That is, when time is short, regardless of chronological age, people concentrate on the most important parts of their lives. For example, in a study of gay HIV positive and negative men, increasing closeness to the end of life, as categorized by health (HIV negative, HIV positive-asymptomatic, and HIV positive- symptomatic), was associated with greater affect in the mental representations of social partners (Carstensen & Fredrickson, 1998). These results suggest that closeness to death is associated with a desire to spend time with people who elicit positive affect, such as family members or close friends. Given the multiple aspects of PTG, it is clear that there is overlap across each of them. In addition, their interdependence may be beneficial by yielding a multiplicative effect. For example, if a trauma survivor re-prioritizes her life goals to appreciate life more, positive family and friend relationships and self-reliance may soon follow. However, a person may report growth in one area while not experiencing growth in another. A person’s relationships may strengthen while other areas (e.g., spirituality, philosophy o f life) may remain static or actually weaken. Finally, it should be noted that the relationship between PTG and adaptation is most likely bi-directional or reciprocal. That is, PTG may facilitate adaptive coping while adaptive coping could simultaneously influence PTG. Therefore, PTG is considered both a process and an outcome (Tedeschi, Park, & Calhoun, 1998). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. increase one’s spirituality whereas the opposite may occur if one feels betrayed by God through the trauma (see Pargament, 1997 for a review o f religion and coping). Dispositions Given previous research regarding adaptation in general and PTG specifically, it appears that PTG may depend on certain dispositions. Particularly, since optimism, locus of control, self efficacy, hardiness, and resiliency have been associated with successful coping, they are also assumed to be associated with PTG (Tedeschi & Calhoun, 1995). In particular, optimism appears to show promise in promoting PTG, although there are mixed findings (see Affleck & Tennen, 1996 for a brief review). A recent discussion has suggested the possibility of item overlap between the predictor (optimism) and criterion (PTG) (Tennen & Affleck, 1998). Therefore, research using a revised measure of optimism (e.g., the revised Life Orientation Test, LOT-R.), which does not contain items concerning benefit finding (e.g., “I’m a believer that ‘every cloud has a silver lining’”), is needed to (dis)confirm this relationship. Psychological Distress As mentioned in the previous section, the distress caused by one’s trauma is considered a catalyst for PTG. Since a trauma is the beginning o f a difficult time requiring ongoing adjustment, growth can occur simultaneously with negative sequela, such as depression. The absence o f depression or distress does not mean PTG has occurred, nor is the absence o f distress necessary for growth to occur. Thus, a large association between PTG and depression/distress is not expected. 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, PTG is assumed to aid in adjustment as indicated by higher levels o f mental health indicators. For example, one study found PTG to predict lower levels o f distress over time, adjusting for baseline levels o f distress (Davis, Nolen- Hoeksema, & Larson, 1998). In fact, this inverse relationship between PTG and distress grew stronger over time. These results suggest that the process o f PTG has a beneficial impact on mental health adjustment overtime. It is too early to discern which person-level and social-level attributes are the most important in facilitating PTG. For example, age and gender may play a larger role than religiousness, although this may vary among both by type o f trauma and PTG outcome. PTG and Adjustment to HIV Although HIV is becoming more of a chronic disease, it remains important to study any factors that are associated with adjustment to disease, the maintenance or initiation of health promoting behavior, and other various mental and physical health or quality of life indicators. A major question is whether PTG should be appreciated as a positive outcome by itself, or viewed as a means to some other end. To date, a number o f studies indicate that there is an association between PTG and various health indicators. For example, among military veterans perceptions o f the positive effects o f military service was negatively related to posttraumatic stress disorder levels (Aldwin et al., 1994). Finding something positive in the loss o f a loved one, such as growth in character or relationship improvement, has been related to lower levels o f post-loss distress, after adjusting for pre-loss distress (Davis et al., 1998). 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In a study of heart attack victims, those who found benefits (e.g., change in philosophy of life) 7 weeks after their first attack were less likely to have another attack and had lower levels of morbidity 8 years later (Affleck, Tennen, Croog, & Levine, 1987). Among people whose homes were damaged or destroyed by a fire, focusing on the positive aspects was associated with lower levels of symptom reporting (Thompson, 1985). Another study examined levels of cortisol reactivity among women exposed to a lab stress, such as solving a difficult math problem (Epel, McEwen, & Ickovics, 1998). Prolonged elevations of cortisol can break down the body’s resources and speed up the disease process. High levels of PTG were associated with lower levels of cortisol reactivity, suggesting that PTG is related to a resilient physiological response towards stress. To explain this finding the authors suggested that those with high PTG may have coped with the lab stressors more efficiently because they adapt to life stressors (in general) more easily or that they had experienced a form of physiological eustress, a toughening up from their past traumas. However, it is unclear whether PTG influences these health outcomes indirectly (e.g., though health behaviors), because these relationships were not examined. A number of studies have explicitly examined PTG among HIV positive populations. A qualitative study has described finding meaning (positive changes) through HIV to include increases in feelings of control, a sense of community belonging or membership, a ‘here and now’ focus, belief in an afterlife, and altruistic behavior (Schwartzberg, 1994). A second study found perceived benefits among 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. women living with HIV/AIDS to include positive changes in health behaviors (Siegel & Schrimshaw, 2001). A third study discovered that finding meaning (i.e., a newfound respect for life and a commitment to ‘live every day to the fullest’) after the death o f a loved one to be linked with greater immune system functioning (less rapid declines in CD4 T cell levels) and lower rates of AIDS-related mortality (Bower, Kemeny, Taylor, & Fahey, 1998). Because PTG is an indicator of higher levels of functioning and well-being, it is logical to consider additional behavioral or physical benefits that may coincide with PTG. Dispositional Optimism and Pessimism Optimism is a disposition that plays an important role in the resiliency that humans show during times o f psychological and physical challenge. Although optimism has a number of definitions, and thus different measurements, an optimist is generally thought of as having a positive outlook on life and seeing the glass as half full rather than half empty. The majority of scientific research in this area has measured dispositional optimism with the Life Orientation Test (LOT), where optimism is defined as the presence o f generalized positive expectations (Scheier & Carver, 1985). These expectations indicate that optimists generally expect good, rather than bad, things will happen to them. This definition stems from the processes that underlie the self-regulation of behavior (e.g., Carver & Scheier, 1981) and the notion that behaviors are greatly influenced by the expectations about the consequences of those behaviors (e.g., self-efficacy, Bandura, 1986). That is, when people view a desired outcome as attainable, they will attempt to achieve this 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. outcome even if it is difficult and there are many obstacles in their way. If the outcome is not seen as attainable, then people will not attempt or will withdraw their effort towards obtaining it. Within this view, expectancies are a major influence of whether one continues to strive towards a goal or gives up (Scheier & Carver, 1992). Because optimists expect positive outcomes, they are expected to be more likely than pessimists to believe that they can overcome the obstacles they face. However, rather than examine these expectancies as situation-specific, this definition allows optimism to be characterized by general expectancies, and thus to have global, although distal, effects on behavior. That is, since expectancy judgements can range from very general (e.g., “I expect good things will happen to me”) to very specific (e.g., “I expect to enjoy exercising this morning”), general expectancies are expected to affect a wider range of outcomes whereas specific expectancies are expected to have their greatest influence on specific outcomes (Scheier & Carver, 1987). Thus, the LOT consists of eight items tapping generalized expectancies, half worded negatively (“I hardly ever expect things to go my way”) and half phrased positively (“I’m always optimistic about my future”). Measurement Issues Is it optimism? The most notable concern with the construct of optimism is that the effects attributed to optimism may be due to shared variance with trait anxiety/neuroticism (Smith, Pope, Rhodewalt, & Poulton, 1989). High correlations between optimism and trait anxiety/neuroticism have been reported, and the significant relationship 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between optimism and physical symptoms have been eliminated by controlling for the effects o f trait anxiety (e.g., Scheier, Carver, & Bridges, 1994, study I; Smith et al., 1989). However, many studies show significant associations between optimism and psychological and physical well-being even after adjusting for neuroticism (e.g., Mroczek, Spiro, Aldwin, Ozer, & Bosse, 1993; Raikkonen, Matthews, Flory, Owens, & Gump, 1999). Likewise, low optimism or high pessimism is not necessarily another measure for depression. Although moderate inverse correlations between optimism and depression have been observed (e.g., r=-49 with the Beck’s Depression Inventory), this does not indicate that these constructs are completely redundant. This is because they are qualitatively different. Depression is a multifaceted construct whereas optimism and pessimism are not. Thus, it is reasonable to assume that a portion of depression overlaps with pessimism because measures o f depressive symptoms (e.g., CES-D) contain items concerning future orientation (e.g., “I felt hopeful about the future”). However, depression encompasses a myriad of other components, including mood, worry, self-doubt, social withdrawal, loss of energy, indecisiveness, etc. The items from the LOT solely consider generalized expectations (“in uncertain times, I usually expect the best”), not current affect, self-concept, or specific symptoms. Although the LOT does not assess affect or motivation, they are theoretically related (e.g., pessimism is considered a risk factor for depression, although it is possible for an optimist to display depressive symptoms). Thus, outcome expectancies are much more specific than depression. 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Secondly, pessimism/optimism and depression show important quantitative differences, indicating both convergent and discriminant validity. That is, they are related yet are associated with other variables independent of one another (suggesting independent, unique effects). In development, the LOT items tended to load by themselves in factor analyses examining the items from the LOT with items from locus of control, self-esteem, hopelessness, and depression scales (Scheier & Carver, 1985). Furthermore, studies find optimism to predict lower levels o f depression over time after adjusting for baseline levels o f depression (e.g., Bromberger & Matthews, 1996; Epping-Jordan et al., 1999; Nolen-Hoeksema, Girgus, & Seligman, 1992). In addition, a number of studies find non-significant associations between optimism and depression (e.g., Marshall & Lang, 1990; Scheier et al., 1989; Scheier et al., 1994; Study 1), although these studies controlled for various demographics and/or other psychological variables. If these were redundant constructs, these results would not be possible. Finally, studies find optimism to predict a number of outcomes after controlling for depression (e.g., rehospitalization, Scheier et al., 1999; cancer mortality, Schulz, Bookwala, Knapp, Scheier, & Williamson, 1996). Because optimism significantly predicts both depression and other outcomes after controlling for baseline depression in longitudinal studies, it is clear that optimism and depression, although importantly related, are not redundant constructs. Thus, although there are important confounds to take into consideration, optimism has remained as an important independent construct. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Is pessimism more important than optimism? While the LOT was originally developed to be a unidimensional measure of optimism and the majority of previous research utilized the LOT as a single optimism scale, there is strong evidence supporting the use of optimism and pessimism subscales. Factor analyses have led a number of researchers to examine pessimism and optimism separately, yielding different effects for each subscale (e.g., Lai, 1994; Marshall, Wortman, Kusulas, Hervig, & Vickers, 1992; Raikkonenet al., 1999; Robinson-Whelen, Kim, MacCallum, & Kiecolt-Glaser, 1997). Furthermore, these studies suggest that the pessimism subscale is a more robust predictor of health outcomes than the optimism subscale. These studies suggest that optimism and pessimism are two unipolar dimensions. Thus, a lack of pessimism is distinguishable from the presence of optimism. If pessimists are not as active in pursuing healthy goals, as are optimists, the result can lead towards faster disease progression or inhibited recovery, confirming pessimism as a risk factor. Although their conceptual differences remain murky, because studies find optimism and pessimism to be empirically distinct, future research will benefit by utilizing similar subscale analyses to further explore whether thinking pessimistically is more important than thinking optimistically, or whether optimism reduces the beneficial effects o f pessimism. Process o f Optimism/Pessimism A construct such as optimism/pessimism can theoretically influence health three different ways: through a direct effect, through mediating effects, and through 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. moderating effects. To better understand the processes underlying optimism/pessimism, each of these pathways will be examined separately. Direct effects of optimism/pessimism Theoretical reviews (Scheier & Carver, 1987, 1992) and a meta-analysis (Andersson, 1996) find wide support for a positive relationship between optimism and various health outcomes for both healthy and diseased populations. In a meta analysis of the LOT consisting of 30 studies, the weighted correlation between optimism and symptom reporting was -0.23 (Andersson, 1996). This relationship is reported among both subjective self-reports of well-being as well as more concrete behavioral outcomes. This relationship is also found across studies utilizing different subjects and a wide range of methods and outcomes, with most controlling for important covariates (e.g., neuroticism, baseline health status, stress levels, etc.). For example, optimism is related to less symptom reporting in college undergraduates in both cross-sectional (O’Brien, VanEgeren, & Mumby, 1995; Sumi, 1997) and longitudinal studies (Chang, 1996; Lai, 1994; Scheier & Carver, 1985). Other studies find this relationship among middle aged men (Mroczek et al., 1993) and women (Thomas, 1995), as well as among family caregivers and non caregivers (Robinson-Whelen et al., 1997). Likewise, in a 35 year study of Harvard graduates, pessimism (explanatory style) at age 35 was predictive of poorer physical health (indicated by physician examinations) at ages 45 through 60 (Peterson, Seligman, & Vaillant, 1988). Thus, recent findings that pessimism increases the risk o f mortality is not surprising (Maruta, Colligan, Malinchoc, & Offord, 2000). 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Among pregnant women, a personal resources factor (which included optimism) was a significant predictor of higher birth weights (Rini, Dunkel-Schetter, Wadhwa, & Sandman, 1999). The influence of optimism is also found among diseased populations. For example, optimism is associated with faster physical recovery (Scheier et al., 1989) and fewer rehospitalizations (Scheier et al., 1999) after coronary artery bypass surgery. Likewise, a cognitive adaptation index (which included measures o f self esteem, control, and optimism) predicted fewer coronary events and greater adjustment after angioplasty (Helgeson, 1999; Helgeson & Fritz, 1999). Among patients suffering from arthritis, pessimism is associated with greater disease impact, indicated by arthritis-related impairment and pain (Brenes, Rapp, & Miller, 1999; Smith, Wallston, & Dwyer, 1995; Tennen, Affleck, Urrows, Higgins, & Mendola, 1992). Evidence suggests that the influence o f optimism may vary depending on the disease. Unlike heart disease and arthritis populations, few physical benefits from optimism are found among cancer patients. Among cancer patients outcomes with null findings include symptoms (hearing loss, Andersson, 1999), disruptions in recreational activities (Johnson, 1996), and fatigue (Smets, Visser, Garssen, Frijda, Oosterveld, & Haes, 1998). However, one study found pessimism to be a risk factor for cancer mortality among younger recurrent cancer patients, suggesting that age may be an important moderator (Schulz et al., 1996). Although these differences between diseases may depend on study designs (e.g., statistical power), they may 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. also be an indication of the control that one has over the trajectory of one’s disease. That is, active behavioral coping strategies, which are associated with optimism, may be more effective with diseases like arthritis and heart disease where behavioral changes have a greater influence on disease progression than diseases such as breast cancer. Taken together, the literature supports a positive relationship between optimism (or lack of pessimism) and health. However, it is dangerous to assume a direct causal link between optimism and health, as caution is needed whenever the relationship between health and psychological variables are examined. These associations must be integrated with existing known predictors of health outcomes. Thus, the underlying mechanisms that may account for this relationship need to be carefully examined. Mediators: Pathways Through which Optimists Stay Healthy The literature indicates three general pathways through which optimism operates: mental health (e.g., lower levels of psychological distress), health behavior (e.g., through active coping strategies), and physiological functioning (e.g., immune functioning). These pathways are not mutually exclusive. For example, improving mental health through the reduction overall psychological distress/depression is an important goal because o f it’s relationship with a number o f important health outcomes. First, evidence supports a relationship between psychological distress (and coping) and the immune system. For example, distress is related to susceptibility to colds (Cohen, Tyrrell, & Smith, 1991) and to a response towards a 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. viral vaccine (Glaser, Kiecolt-Glaser, Bonneau, Malarkey, & Hughes, 1992). Indeed, research on the enhancement of immune function through different behavioral strategies (e.g., stress reduction) shows promise (Chesney & Folkman, 1994; Kiecolt-Glaser & Glaser, 1992). Second, distress is related to important health behaviors, such as adherence to treatment, suggesting that compliance may mediate the distress to health relationship (DiMatteo et al., 1993). For example, individuals who are distressed may be less likely to adhere to their treatment regimen, which could affect disease progression. Third, psychological distress may lead to changes in important health behaviors. For example, distressed patients may smoke and drink alcohol more than, and be less likely to maintain or initiate a proper diet or exercise regimen than non-distressed patients (e.g., DiMatteo et al., 1993). Thus, although examined separately, each proposed mediator is assumed to have additional health benefits through their respective influence on one another. Mental health. Optimism and pessimism are associated with a variety o f measures o f mental health and well-being (e.g., depression, stress, anxiety). For example, a meta-analysis of the LOT consisting o f 30 studies, found the weighted correlation between optimism and negative affect to be -0.43 and between optimism and the Beck Depression Inventory (using 5 studies) to be -0.45 (Andersson, 1996). Importantly, these relationship are found among longitudinal studies after adjusting for baseline factors among both healthy (e.g., Chang, 1996; Chang & Bridewell, 1988; Hull & Mendolia 1991; Segerstrom, Taylor, Kemeny, & Fahey, 1998; Sumi, 1997; Raikkonen et al., 1999) and diseased (e.g., Tennen et al., 1992) populations. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, the association with depression has become non-significant after controlling for a variety of covariates, such as self-mastery, trait anxiety, neuroticism, and self-esteem (Marshall & Lang, 1990; Scheier et al., 1994, study 1). Among cancer patients, optimism is repeatedly related to higher levels of mental functioning. For example, optimism is related to greater levels o f positive mood in prostate cancer patients during and after receiving radiation therapy (Johnson, 1996) and lower levels o f negative mood among long-term survivors of bone marrow transplantation (Curbow, Somerfield, Baker, Wingard, & Legro, 1993). Optimism is inversely related to depression and anxiety among both prostate (Bjorck, Hopp, & Jones, 1999) and breast cancer (Epping-Jordan et al., 1999) patients. Among recurrent cancer patients seeking radiation therapy, optimism is related to less depression and perceived stress, and related to greater psychological well-being/affect balance (Schulz et al., 1995). Among early stage breast cancer patients, optimism is inversely related to distress and positively related to well-being after surgery (Carver et al., 1993; 1994). Studies also indicate that optimism/pessimism can influence mental and physical health through stress reduction. This is not surprising as many studies have found positive relationships between low optimism (or high pessimism) and levels of distress or anxiety (e.g., Litt, Tennen, Affleck, & Klock, 1992; Mroczek et al., 1993; O’Brien et al., 1995; Robinson-Whelen et al., 1997). Importantly, a number of studies have indicated an inverse association between optimism and psychological distress among men with HIV or AIDS (e.g., Kang, Bjorck, & Watts, 1997; Taylor et 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. al., 1992). This is particularly important in relation to HTV where high levels o f stress have been associated with disease progression (e.g., Leserman et al., 1999). Behavioral health. Optimism has positive associations with a variety of healthy behaviors. For example, optimism has been related to lower levels alcohol, cigarette, and other substance use and intent to use in the future (Carvajal, Evans, & Nash, 2000; Carvajal, Clair, Nash, & Evans, 1998) and higher levels o f exercise and diet behaviors (Robbins, Spence, & Clark, 1991). There is also evidence suggesting that optimists are more likely to pay attention to health-risk information (Aspinwall & Brunhart, 1996). However, a number of studies have found no association between optimism and various health behaviors, such as alcohol use, exercise, and sleep patterns (e.g., O’Brien et al., 1995; Segerstrom et al., 1998). Once notified o f a potential health problem, optimists appear more likely to act. For example, in a study on skin cancer, optimism predicted subsequent compliance with medical follow-up (Friedman, Webb, Bruce, Weinberg, & Cooper, 1995). That is, of those subjects who were identified as having suspicious lesions, optimists were more likely to have had the lesion examined by a physician. Optimism also predicted higher sunscreen use intentions. Similarly, optimism has been associated with medical follow-up after discovery o f breast cancer symptoms (Lauver, 1994). A subsequent study with the same participants found optimism to predict less care seeking delay (number o f days between finding a symptom and contacting the health care system), although this relationship became nonsignificant after adjusting for race, education, and occupation (Lauver & Tak, 1995). A similar 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lack o f significance with medical follow-up has been found after receiving an abnormal papanicolaou test result (Lauver & Rubin, 1990). In an 18 week study of patients enrolled in a cardiac rehabilitation program, optimism predicted reaching the goals of lowering body fat, saturated fat, global coronary risk and increasing aerobic capacity (Shepperd, Maroto, & Pbert, 1996). Further analyses revealed that coping, specifically problem focused coping strategies, mediated these relationships. Thus, once made aware of a potential health problem optimists appear quick to respond, although it may depend on the circumstances (e.g., abnormal pap vs. abnormal SSE). Nevertheless, it is too early to declare a strong relationship between optimism and adherence to medical protocols. However, one study has found a significant relationship between optimism and adherence to antiretroviral medications among HIV positive men (Chesney, Folkman, & Chambers, 1996). Since less than 100% compliance may compromise one’s viral load, this finding is particularly important. It is reasonable to conclude that optimists work harder and sooner at achieving positive outcomes than do pessimists because optimists expect them (Scheier & Carver, 1985, 1987). Likewise since pessimists are more likely to act passively during difficult times, behavior should be a major link between optimism and health. However, most of the empirical evidence supports behavioral pathways that are reactive (such as compliance and coping), although there is some evidence supporting links with proactive behaviors (such as diet, exercise, and the avoidance of substance use). Thus, effective coping strategies, how one responds to alleviate 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. stressful situations, are more likely to be utilized by optimists. Nevertheless, because it is too early to conclude that optimism has an overwhelming effect on health promoting behaviors, research is needed to further examine the possibility. There is an indication that optimism may be associated with lower levels of realism in HIV populations. For example, in a sample of gay HIV+ and HIV- men, optimism was associated with greater perceived control (primarily for HIV+ subjects) and less AIDS-related worries & concerns, and perceived risk of developing AIDS (Taylor et al., 1992). However, because optimism was not related to unhealthy behaviors (e.g., poor diet, lack of exercise, unsafe sex), the notion that optimist’s may engage in unhealthy behaviors due to denial or apathy stemming from their positive outlook was not supported. Nevertheless, the possibility that optimism fosters unrealistic health assessments that may influence apathy towards health needs to be examined. Coping. The stability of the relationships between optimism/pessimism and various coping strategies has been reinforced by similar findings across both time and various studies. For example, optimism has been related to more active coping, planning, positive reinterpretation and growth, seeking social support and less behavioral disengagement across time (Billingsley, Waehler, & Hardin, 1993). Optimism is positively associated with planning, seeking social support, positive reinterpretation, and turning to religion coping strategies (e.g., Scheier et al., 1994, study 1). 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Repeatedly, optimism has been either unrelated or inversely related to negative coping strategies. These results have been found for strategies characterizing avoidant or emotion-focused coping, such as behavioral and mental disengagement, alcohol use, denial, or restraint (e.g., Billingsley, et al., 1993; Segerstrom et al., 1998; Scheier et al., 1987, study 2, 1994, study 1; Strutton & Lumpkin, 1992). There is also research indicating these same relationships, greater active behavioral and less avoidance coping strategies, are present among diseased populations, including cancer (Friedman et al., 1992), arthritis (Long & Sangster, 1993) and heart disease (Scheier et al., 1989) patients. Among gay HIV+ and HIV- men, optimism has been associated with greater positive attitude coping and less fatalism/self-blame/escape-avoidance coping (Taylor et al., 1992). These relationships are important because adaptive coping strategies have been associated with lower levels of depression and mood disturbance (e.g., Namir, Wolcott, Fawzy, & Alumbaugh, 1987), increased immune functioning (Goodkin et al., 1992), and higher levels o f survival (e.g., Soloman, Temoshok, OTeary, & Zich, 1987) among people living with HTV/AIDS. Physiology. In relation to physical health, physiological indicators (e.g., immune functioning) are the most important (i.e., proximal) mechanisms to take into account. Yet compared to the number of empirical studies supporting the other pathways, there is a definite need for more research in this area. Nevertheless, there is evidence confirming positive effects of optimism on a number of biological 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. markers. For example, in a 6-month study of patients who underwent coronary artery bypass surgery, optimists showed fewer new Q-waves on their EKGs and release o f AST than pessimists during coronary artery bypass surgery, both of which are signs of myocardial infarction (Scheier et al., 1989). (Although the authors did note that the base rate for myocardial infarctions in studies of this type is low.) Pessimists have shown significantly higher average diastolic blood pressure (DBP) than optimists, but no difference for systolic blood pressure (SBP) (Raikkonen et al., 1999). Likewise, in an experimental study involving a stress task, pessimists displayed greater DBP reactivity to the task than did optimists (Williams, Riels, & Roper, 1990). No differences in SPB or heart rate reactivity was observed. These results suggest that pessimism may be associated with hypervigilance. That is, pessimists may display greater cardiovascular reactivity than optimists because they are more vulnerable to stressful events. In a 10 week study of 1st year law-school students, optimism was unrelated to a number of immune parameters (both helper T cells and natural killer cells), after adjusting for baseline levels (Segerstrom et al., 1998). However, there was a positive association with situational optimism, indicating that specific/situational optimism may be more robust in predicting immune outcomes. However, in a study of older adults, pessimistic explanatory style was associated with a lower T4/T8 ratio and lower T-lymphocyte response to mitogen challenge (Kamen-Siegel, Rodin, Seligman, & Dwyer, 1991). 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In a 3-month prospective study of 39 healthy women, optimism was found to moderate the relationship between stress and a number of immune parameters (Cohen et al., 1999). This study examined natural killer (NK) cells and two subsets of T cells, the Helper-Inducer T cells (CD4) and subsets of cytotoxic T cells (CD8). For CD8+CD1 lb+, optimists levels were not related to weekly levels o f acute stress. Pessimists levels were inversely related to stress. For CD8+CDI lb- cell percentages and NK cell cytotoxicity, optimists experiencing high weekly levels of persistent stress showed lower levels of immunity than optimists experiencing low levels of persistent stress. Pessimists level o f immunity had a positive relationship with persistent stress levels for CD8+CD1 lb- and no relationship with NK cell cytotoxicity. Thus, it appears that optimism buffers the effect o f acute stress, but enhances the effect of chronic stress on these immune parameters. The authors note that this result is in line with the notion that optimists may have difficulties in situations of prolonged difficulties that are inconsistent with their expectations. This is also consistent with the findings from a study in which optimists showed a reduction in immunoreactivity (NK activity) subsequent to uncontrollable stress (Sieber et al., 1992). This reduction was only present immediately after uncontrollable stress exposure (20 minutes), not later (80 minutes, 24 hours, or 72 hours later). Thus, control appears to play an important role in relation to immune responses among optimists. Few studies have examined optimism in relation to disease progression and most do not find associations. For example, two recent studies found no association 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between dispositional optimism and CD4 count over time (Anderson, 2000; Tomakowsky, Lumley, Markowitz, and Frank, 2001). Because these studies had small samples (both n’s<100) they may not have had enough statistical power to determine whether such a relationship existed. In a study o f African-American women co-infected with HIV and human papillomavirus (HPV), pessimism (Millon Behavioral Health Inventory) was associated with lower levels o f natural killer cell (NKCC) percentage and marginally (p < l0 ) associated with lower levels of T cytotoxic/suppressor cell (CD8+CD3+) percentages (Byrnes et al., 1998). This was after adjusting for HPV status, behavioral/lifestyle factors (e.g., sexual, drug behaviors), and stressful negative life events. Moderating Effects: Optimism as a Stress Buffer A number of studies have examined optimism under the stress-buffering model. Unlike the main effect model that proposes that optimism is beneficial in maintaining health regardless of whether or not an individual is under stress, the stress-buffering model proposes that optimism “buffers” an individual from the harmful effects o f stressful life events and daily hassles. That is, those who are under high stress will experience negative health outcomes in the absence of optimism; whereas in the presence o f optimism fewer, if any, negative outcomes will be experienced. Thus, the model suggests there is an interaction between stressful events and optimism as these relate to health outcomes. Across a variety o f studies, optimism has shown to buffer the negative effects of various stressful events. In particular, optimism has been found to buffer the 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. relationships between stress and: fatigue/anxiety (Williams et al., 1990), depression (Bromberger & Matthews, 1996), executive burnout (Fry, 1995), and symptom reporting (Lai, 1995). That is, under high stress, those who scored low in optimism had higher levels of negative outcomes than those who scored high in optimism. Thus, the evidence indicates that optimism is not only related to lower levels of stress and symptomatology, but also that when optimists are under high levels of stress, it is less likely to lead towards poor physical or mental health. As noted earlier, this is important in relation to HIV where high levels of stress has been associated with disease progression (e.g., Leserman et al., 1999). Summary In summary, optimism and pessimism have important influences on health outcomes for both healthy and diseased populations. When healthy, optimism is related to fewer negative health outcomes, especially during times o f stress. When diseased, under definite health-related stress, optimists have higher levels of various adjustment indicators. These relationships are found among studies utilizing both subjective self-reports and objective physiological outcomes. In addition, it is evident that many of the same effects (i.e., mediators) and illness avoiding techniques that characterize optimists when healthy are also in effect and utilized to get better when they are ill. Given the wide array o f stressors that optimists have successfully managed (ranging from beginning college and childbirth to bypass surgery and cancer), it is clear that optimism aids in the ongoing adjustment to life’s difficulties. 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, studies examining more direct health outcomes (i.e. physiology, preventive behaviors) are needed. The strength o f these relationships, particularly among those living with HIV, is unclear. Longitudinal designs are needed to understand the longer-term influence and benefits that optimism has on overall health status, particularly through these pathways. Thus, testing for additional mediators of the optimism-^ health relationship is important. The studies that have tested specifically for mediation provide important evidence for how optimism operates on one’s health status. Although the stress and adaptive coping pathways appear to be the most robust, there is a clear need for more psychoneuroimmunological studies, as this is an area where studies are lacking. Positive Beliefs and HIV HIV, compared with other diseases, provides a superior model for studying the influence o f positive beliefs on health. There are several reasons for this benefit (as noted by Taylor, Kemeny, Reed, Bower, & Gruenewald, 2000). First, individuals with HIV can be identified while asymptomatic allowing for the following of disease course over time. This also allows one to examine possible mediators of the positive beliefs-^disease course relationship. For example, optimism may influence a positive change in health habits, which may subsequently prevent the onset of symptoms. Second, there are known covariates that can be statistically controlled for in analyses to avoid potential confounding. For HTV, these include demographics (e.g. age, SES) as well as various health behaviors (e.g., alcohol/drug use). A third 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. advantage is that HIV has important clinical markers associated with infection, including viral load, CD4 counts, symptom appearance, and AIDS diagnosis. Unfortunately, aside from the studies already mentioned, few investigators have examined the relationship between positive beliefs and HIV, although those who have find interesting results. In a study of 78 gay men with AIDS, the relationship between mortality and various outcome expectancies and coping strategies was examined (Reed, Kemeny, Taylor, Wang, & Visscher, 1994). The men in this study were mostly of high socioeconomic status (SES), White, and had been diagnosed with AIDS for roughly one year. Although dispositional optimism was not related to survival time, a method of coping termed realistic acceptance was. This construct consisted of AIDS specific fatalism and resignation (e.g., “I tried to accept what might happen,” “I prepare myself for the worst”), a sense of pessimism. Furthermore, this relationship was independent of various covariates including CD4/CD8 ratio, health status, distress, adherence, health behaviors/risk factors, and other coping strategies. In a separate study, the combination o f negative HIV- specific expectancies (low levels o f control, confidence, and optimism in regard to the future course of one’s HIV-related illness) and suffering from AIDS-related bereavement was predictive o f symptom onset in a sample o f asymptomatic HIV- positive gay men (Reed, Kemeny, Taylor, & Visscher, 1999). Again, this relationship held after controlling for a myriad o f covariates. These studies suggest that negative expectations regarding one’s illness is related to disease progression, indicating that psychosocial resources protecting 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. against negative expectations can have salutary effects (Taylor et al., 2000). Indeed, evidence suggests that PTG may be one of these resources. A study examining 40 HIV positive men who experienced the loss of a close friend or partner to AIDS, discovered that finding meaning (i.e., a newfound respect for life and a commitment to ‘live every day to the fullest’) after the death was linked to greater immune system functioning (less rapid declines in CD4 T cell levels) and lower rates of AIDS- related mortality (Bower et al., 1998). Given these results, the many studies examining the beneficial effects o f optimism among other diseased populations, and the promise of PTG and the notion of finding meaning, examining the specificity of these relationships among people living with HIV/AIDS is a logical next step. 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PRESENT STUDY It is important to explore the benefits o f optimism/pessimism and PTG among illnesses such as HIV, where complete recovery is not possible. Once diagnosed, an individual can experience tremendous amounts of stress. In addition to dealing with the notion of one’s impending death, because treatment advances have increased survival rates, sources of stress also include factors relating to living. Adjustment to these stressors can influence a patient’s psychological, behavioral, and immune responses. Thus, reactions towards the traumatic event of diagnosis and subsequent adjustment are important in managing disease progression over time. Research is needed to identify the positive changes and strengths in people living with HIV/AIDS and to further determine the subtle aspects that influence both the processes (i.e., mediators) and the different outcomes o f PTG, optimism, and pessimism. Examining the relationship between psychological variables and disease progression among lower SES populations is important. Previous studies examining psychological variables within HTV populations include mostly White, affluent, and well-educated men who had good access to medical care. If similar effects are found when these important factors are not at their highest (e.g., quality care, nutrition, etc.), then psychological variables may not necessarily be completely subsumed by biological variables. On the contrary, they may become more relevant when other resources are not available. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Previous investigations clearly indicate that PTG is possible after the diagnosis of a life-threatening illness. The correlates of posttraumatic growth appear robust, although they show some variation across different populations. Given that research on PTG is still in its infancy, the range o f situations in which PTG occurs is unknown. The first major question in the present study is whether PTG can occur among a sample of patients diagnosed with HIV. Although recent combination drug therapies show promise in turning HIV into a chronic disease, diagnosis is still considered a life and death issue (i.e., a severe traumatic event). PTG is expected to be present among this sample. Second, there is little data available which examines the possibility of experiencing health benefits secondary to experiencing PTG. In addition, optimism and pessimism have a history of predicting various health outcomes. The second question in the present study is whether there are physical health benefits from optimism/pessimism or experiencing PTG. In this study, PTG and optimism/pessimism are expected to influence disease progression. Disease progression was measured using HIV RNA viral load (a measure of the quantity of HIV in the blood, high numbers suggest a high risk o f disease progression) and CD4 count (a measure of T-helper cells, low numbers suggest greater damage to the immune system). If there are relationships between these variables, then health behaviors (e.g., diet, exercise, etc.) and metal health (i.e., depression) will be examined as possible mediators. For example, if pessimism is positively associated 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with alcohol use, and pessimism is inversely associated with CD4 counts over time, then alcohol use may account for some or all o f this relationship. Third, does optimism/pessimism or experiencing PTG influence mental health in this population? PTG and optimism are expected to positively influence and pessimism is expected to negatively influence mental health, as indicated by lower levels of depression. It is possible that the relationship between psychological factors and physical/mental health outcomes is moderated by certain conditions, such as SES or baseline health status. For example, there is the growing evidence that expectancies interact with bereavement to predict disease progression (Bower et al., 1998). Thus, it may be that when stress increases (e.g., low SES, poor health) psychological factors become more potent. This notion is consistent with the stress-buffering role of protective factors. Alternatively, psychological factors may also have a larger influence earlier in the disease course when biological influences may have relatively less impact (Scheier & Bridges, 1998). However, because cofactors of disease progression exert their influence cumulatively over time, there may be greater differences in disease status with increasing time since infection. Simply controlling for disease status (and thus the heterogeneity of the participants in health status) may eliminate any cofactor-related differences in disease status that emerge during the time from infection to data collection. In order to examine these possibilities appropriate interaction terms were explored (SES by optimism, baseline disease status by optimism, time since diagnosis by optimism, etc.). 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fourth, although a number of correlates appear to be consistent in their relationship with PTG, it is possible that these relationships vary across both time and population. Thus, the fourth question in this study is whether these correlates of PTG are predictive of PTG over time among this sample. Regarding demographics, it is expected that females and higher SES participants will have more PTG than their respective counterparts. Since the process of PTG appears to take time in order to allow for proper rumination of existential issues, those who have been recently diagnosed are expected to have less PTG than those who have had ample time to deal with their diagnoses. Those who score high in optimism and low in pessimism are expected to have more PTG than those who score low in optimism or high in pessimism, respectively. Although optimism and pessimism are assumed to be somewhat stable, recent research indicates that levels can change over time and be influenced through interventions (e.g., Mann, 2001). Thus, changes in optimism/pessimism will also be explored. These analyses will also determine whether there are any common cofactors of these constructs. Fifth, are PTG and optimism/pessimism associated with other important factors related to HTV? Specifically are they related to ART adherence and medicine side effect reporting? Study Hypotheses Specific hypothesis for this study are grouped under the 5 aforementioned questions: 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Can PTG occur in a sample of patients diagnosed with HIV? (i .e.. Are there identifiable strengths and positive changes in people living with HTV?") HI. PTG is hypothesized to be present among this sample. Because responses to PTG items can range from (1) highly negative change to (5) highly positive change (neutral scoring 3), participants with a mean score of 4 or greater will be considered to have experienced PTG. Are optimism/pessimism or PTG associated with disease status and progression? H2. Pessimism will be positively associated and optimism and PTG will be inversely associated with disease progression. Disease progression is indicated by higher levels o f viral load and/or lower levels of CD4 count over time. If H2 is supported, then depressive symptoms and health behaviors are expected to partially mediate these relationships. In addition, two-way interactions will be explored to determine whether any factor moderates the relationships between disease progression and optimism, pessimism, or PTG. Are optimism/pessimism or PTG associated with mental health? H3. Optimism/PTG will predict lower levels of depressive symptoms and pessimism will predict higher levels of depressive symptoms. However, because psychological distress and PTG can co-occur, the relationship between PTG and depression is hypothesized to be weaker than the relationship between depression and optimism/pessimism. In addition, two-way interactions will be explored to determine whether any factor moderates the relationships between depression and optimism, pessimism, or PTG. 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What are the correlates of PTG. and optimism/pessimism? H4. Correlates and predictors o f PTG. optimism, and pessimism. Correlates of PTG found in prior studies are expected to be associated with PTG in this sample. These hypothesized relationships include: H4a) Optimism is hypothesized to be positively associated, while pessimism is expected to be inversely associated with PTG, H4b) Females are hypothesized to experience greater levels of PTG than males, H4c) Time since diagnosis is hypothesized to be positively associated with PTG, H4d) SES is hypothesized to be positively associated with PTG, and H4e) Religiosity is hypothesized to be positively associated with PTG. These relationships will be examined concurrently as correlates (at baseline) and in a longitudinal model predicting PTG over time. Levels of optimism and pessimism will also be examined concurrently and in longitudinal models to determine whether there are any common cofactors across these constructs. H4f is that PTG and optimism are expected to be positively and pessimism is expected to be inversely associated with health promoting behaviors (diet, exercise, avoiding substance use, etc.). Are PTG and optimism/pessimism associated with ART adherence and medicine side effect reporting? H5. PTG and optimism are expected to be positively associated with adherence while pessimism is expected to be negatively associated with adherence. These relationships will be explored only among those patients taking ART. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H6. Optimism is expected to be inversely associated with and pessimism is expected to be positively associated with side effect reporting and difficulty in dealing with side effects. These analyses will be performed on both total scores as well as on those side effects determined to be more psychosomatic versus biological in origin. There are no hypotheses concerning PTG for side effects. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. METHOD Participant Recruitment At 6 outpatient clinics located in California, trained data collectors screened patients for inclusion in the study. While males were randomly approached for inclusion in the study, all female patients were approached because there are few female relative to male patients. Patients were eligible to participate if they met the following criteria: (a) at least 18 years of age, (b) HIV seropositive (asymptomatic, symptomatic, AIDS), (c) had known o f their HIV status for at least 3 months, (d) English or Spanish speaking, (e) receiving care at the clinic where interviewed, and (f) because the larger study was examining sex behaviors, were sexually active within the last 3 months (this includes anything beyond kissing, such as mutual masturbation or oral sex). Enrollment continued until each clinic had recruited approximately 150 patients. Overall, we approached 2,027 patients. Nine percent (n=187) refused to be screened. O f those screened, 562 were ineligible for at least one o f the following reasons: no sexual activity in the past 3 months (88.1%), not receiving care at the clinic (6.4%), diagnosed within the previous 3 months (6.2%), non English or Spanish speaking (0.7%), and under age (0.2%). O f the 1,278 who were eligible, 886 (69%) enrolled. Those who were eligible but not recruited (n=392) refused to participate due to lack o f time (46.7%), not wanting to be in the study (10.0%), too ill (1.8%), gave some other reason (1.8%), or gave no reason (39.0%). 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Demographic information from screening forms was compared between those enrolled and those who did not enroll yet were eligible to determine whether the recruited sample differed from those who did not agree to participate. Significantly more female patients enrolled than female patients who did not enroll. Significantly more Hispanics and those speaking Spanish enrolled than did not enroll and significantly fewer Whites enrolled than did not enroll (all j j ’s < 05). All multivariable analyses statistically adjusted for clinic status to control for any differences between clinics. Procedure Once eligibility was determined and the participant agreed to enroll in the study, data collectors (trained research assistants at each clinic) obtained informed consent which detailed the nature of the study, the potential benefits and risks of involvement, procedures used to protect and ensure confidentiality, a statement of a right to withdraw from the study, and a description of the compensation that would be received for participating in the study ($10 and $15 for completing the pre- and post-measure, respectively). Participants completed a paper-and-pencil survey in a private room located in the respective clinic. Each participant, and respective survey, was assigned a random number for identification purposes. In order to ensure confidentiality, each clinic maintained these original records and kept them in a locked file cabinet to which only the PI and study coordinator had access. Medical information (e.g., CD4 T cell counts, viral load, etc.) was abstracted from medical charts. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The mean follow-up time for the Time 2 survey was 1.57 years. Multiple attempts were made to contact hard to reach participants either by telephone or mail. Three hundred and three participants could not be contacted after repeated attempts and were considered lost to follow-up. Of these, 27 had died. Participants completed the follow-up survey either in the same manner as the baseline survey (n=456) or, for participants who had limited time, with a brief survey containing fewer items administered either in person or over the telephone (n=T27). Measures Demographics. Self-reported demographics included gender, ethnicity, and income. Income was determined from total household income for the previous year. Medical information. Medical information was assessed with data obtained from medical chart reviews at baseline and follow-up (e.g., CD4/T-cell counts, viral load). Viral load was log transformed for use as a continuous variable. Length of time since HIV diagnosis was obtained through self-report. Religiosity. At follow-up, participants were classified as religious if they indicated that they typically attended religious services at least 1-3 times a month or were a member of a church, synagogue, or other place of worship. Health behaviors. Single items ascertained various health behaviors. Weekly levels of exercising and eating a heaithy diet were made on 4-point scales ranging from never (0) to every day (3). Daily cigarette usage was determined on a 5-point scale ranging from never (0) to more than a pack per day (4). Previous 3-month alcohol use was determined on a 7-point scale ranging from never (1) to every day 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (7). Previous 3-month illicit drug use was determined with a yes/no checklist o f 12 substances (methamphetamines, crack cocaine, etc.) and was binary coded with I equaling use and 0 equaling non-use. Marijuana use was not included due to the use of this substance for medical purposes. Depression. Depression was ascertained with the Center for Epidemiological Studies Depression Scale (CES-D) (Radioff, 1977). The CES-D is a 20-item measure of current (i.e., past week) depressive symptoms (e.g., “I was bothered by things that usually don’t bother me” and “I felt lonely”). Responses were made on a 4-point scale, ranging from I (rarely or none of the time, less than one day) to 4 (most or all of the time, 5-7 days). Consistent with previous research utilizing the CES-D among HIV positive populations (e.g., Burack et al., 1993; Ickovics et al., 2001; Lyketsos et al., 1993), 5 somatic items that are related with symptoms of HIV infection (poor concentration, fatigue, poor appetite, lack of energy, and restless sleep) were removed from this scale. The coefficient alpha for this 15-item scale was .91 and .92 at baseline and follow-up, respectively. PTG. PTG was assessed with an 11-item self-report scale. Items were both developed by the investigator and modified from the Posttraumatic Growth Inventory (PTGI; Tedeschi & Calhoun, 1996). Items used in previous research concerning PTG have been positively worded with responses made on scales ranging from “Not at all” or “I did not experience this change as a result of my crises” to “ extremely” or “I experienced this change to a very great degree as a result of my crises.” Thus, negative responses were not made available. That is, if one 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. experienced a weakening of religious faith, they would only be able to indicate no change. While scale authors explain that PTG measures are designed only to ascertain positive growth, constricting the response scale unnecessarily limits the respondent. Unfortunately, this coding scheme gives equivalent scores to participants who had no change and to those who felt they had changed for the worse. The items for this study were developed to ascertain both positive and negative changes. Thus, responses can range from “Highly negative change” to ’ ’Highly positive change” on a 5-point scale for the following items: appreciation for the value of my own life, priorities about what is important in my life, involvement in things that interest me, my understanding o f spiritual matters, direction for my life, my sense of closeness with others, willingness to express my emotions, handling my difficulties, my compassion for others, my religious faith, and my own inner strength. There were no major problems or confusion from participants in the pilot testing of these items. Similar to previous research on PTG (e.g., Antoni et al., 2001), a factor analysis at Time 1 suggests this measure is appropriate as a unitary scale. Although 2 factors had eigenvalues greater than 1, the eigenvalue of the second value was only 1.10 and all items loaded at or above .59 on the unrotated first factor. The internal reliability for this scale was .91 and .92 at baseline and follow-up respectively. Optimism/pessimism. Consistent with research suggesting that the revised Life Orientation Test (LOT-R; Scheier et al., 1994) consists of two independent dimensions with separate optimism and pessimism subscales (e.g., Robinson- 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Whelen, Kim, MacCallum, & Kiecolt-Glaser, 1997), principal components analyses with varimax rotation clearly supported a two-factor solution. Thus, separate scores were calculated for dispositional optimism and pessimism. The 3-item optimism and pessimism subscales had 4-point response formats ranging from “strongly agree” to “strongly disagree.” Examples o f items include “In uncertain times I usually expect the best” (optimism) and “I rarely count on good things happening to me” (pessimism). Cronbach’s alpha’s were .72 and .70 for the optimism subscale and .74 and .73 for pessimism subscale at baseline and follow-up, respectively. Medication adherence. Previous 7-day levels o f adherence to antiretroviral medication regimens were assessed. Participants answered questions for each antiretroviral medication they were prescribed (e.g., “How many times a day were you told to take this drug?” and “How many pills were you told to take each time?”), including the number of pills missed or skipped in the past week. Seven-day adherence (% of pills taken as prescribed) was determined by dividing the total number of pills missed or skipped in the past week by the total number of pills prescribed per week. Due to skewness, medication adherence was dichotomized at 95% (with those indicating 95% or greater adherence scoring 1 and those indicating less than 95% adherence scoring 0). This cutoff is consistent with previous research that suggests that high levels o f adherence are needed for adequate viral suppression (e.g., Paterson et al., 2000). Side effects. Ten unpleasant physical effects experienced from ART were measured with a yes/no format. These included nausea, diarrhea, rash, headache, 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. insomnia, fatigue, weight redistribution, loss of sex drive, tingling/pain in limbs, and blood in urine. The number of side effects was determined by talcing the sum of the side effects experienced. The degree of difficulty in dealing with each of the 10 possible effects was measured with a 4-point scale ranging from “Not at all difficult” to “Extremely difficult.” A total difficulty score was determined by taking the mean of these items. Each of these side effect scores were also separated into two subscales, one score indicating side effects that are more psychosomatic (5-iems: nausea, diarrhea, insomnia, fatigue, and loss of sex drive) and one score indicating side effects that are more biological in origin (5-items: rash, headache, weight redistribution, tingling limbs, and blood in urine). Statistical analyses Statistical analyses began with descriptive analyses o f the variables and an overview of the PTG variable. The relationships between PTG, optimism, and other study variables were first examined in bivariate analyses (t-tests and anova). These results provided preliminary hypotheses testing. The relationships between PTG, optimism, pessimism and disease status and depression were then examined in a series of multivariable regression analyses. Unless otherwise noted, each of these analyses included clinic status, age, gender, ethnicity, SES, antiretroviral use, CD4 count, viral load, depression, PTG, pessimism, and optimism in the model. These analyses were first performed on the baseline data and then repeated with the follow-up data (adjusting for baseline values). In addition, changes in PTG (from time 1 to time 2) were examined in 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. relation to these outcomes in order to test the hypotheses that those who always had or gain PTG will show greater levels of health over time, while those who lost or never had PTG will show decreases in health outcomes. If PTG, optimism, or pessimism were significantly associated with disease status, potential mediators of these relationships were evaluated in additional multivariable regression models with the potential mediator(s) (depression and health behaviors) included and excluded from the regression equation. For example, depression would mediate the relationship between optimism and CD4 count if (1) optimism and depression are significantly associated with CD4 count, (2) optimism is significantly associated with depression, and (3) the association between optimism and CD4 count becomes non-significant when depression is included in the model (Baron & Kenny, 1986). After examining these multivariable models, interaction terms were created by multiplying each variable by PTG, pessimism, and optimism (e.g., PTG by gender) and were added to each model. These analyses determined whether any variable moderated the relationship between PTG, optimism, pessimism and the outcome variables. To understand the nature of significant interactions, mean differences in outcome variables were examined graphically by dividing predictor variables into groups (e.g., median split). Although examining viral loads, CD4 counts, and depression over time were o f primary interest, exploring changes in PTG, optimism, and pessimism over time 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were also examined. Thus, additional multivariable regression models were used to analyze these relationships. 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS Baseline/Longitudinal Samples Three main samples were used in the analyses, 1 baseline and 2 follow-up samples (Figure 1). From the 886 patients recruited into the study, 51 had missing data on baseline study variables, yielding a baseline analytical sample o f 835. The 2 main follow-up samples used in the longitudinal analyses, medical and psychological, were due to differential rates o f missing data across the outcome variables. The medical follow-up sample contains the 744 participants who had valid viral load and CD4 counts abstracted from medical chart reviews done at follow-up. The psychological follow-up sample contains those who completed the full follow-up questionnaire (which contained items concerning health behaviors, PTG, optimism, and pessimism) vs. the brief questionnaire. O f the 584 patients successfully contacted at follow-up, 457 completed this survey and 23 had missing data on follow-up items, yielding a behavioral follow-up analytical sample of 434. Analyses examining adherence to ART and side effect reporting were examined in a subgroup of participants (n=661) from the baseline sample who reported ART use and did not have missing data. These analyses were only performed at baseline because the larger intervention study included an ART adherence component. Attrition Analyses Baseline characteristics were compared (Chi-square and t-test analyses) between participants in the longitudinal samples and those lost to follow-up 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1. Samples 2,027 patients screened 1,278 eligible 562 were ineligible 187 eligibility unknown / 302 lost to follow-up (n=27 died) 584 completed follow-up questionnaire 127 brief version 457 complete version Medical follow-up sample (n=744 with complete biological data for follow-up analyses of CD4 and viral load) 886 (61%) enrolled Baseline sample (n=835 with complete data) Psychological follow-up sample (n=434 with complete data for follow-up analyses of PTG, optimism, pessimism, and depression) 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table I. Characteristics of the Sample at Baseline (n=835) Variable n (%) Mean (SD) Gender Male 727 (87.1) Female 108 (12.9) Age 38.4 (7.8) Ethnicity White 330 (39.5) Hispanic 307 (36.8) African-American 142 (17.0) Other 56 (6.7) Employment status Full-time 184(22.0) Part-time 153 (18.3) Unemployed 498 (59.6) Education Less than high school 218(26.1) High school (or equiv.) 185 (22.2) More than high school 432(51.7) Previous year gross Income <5,000 156(18.7) 5,000-9,999 271 (32.5) 10,000-14,999 133 (15.9) 15,000-24,999 128(15.3) 25,000+ 147(17.6) Years since FIIV diagnosis 6.4 (4.2) Diagnosed with AIDS 387 (46.4) On ART 664 (79.5) CD4 count 407.6 (273.2) Viral load 60,520.5 (247,808.6) 95% adherent to ART (n=6 6 l) 499 (75.5) Side effects experienced (n=661) 4.15(2.63) Psychosomatic 2.53 (1.64) Biological 1.62(1.31) Difficulty with side effects 2.37(0.61) (n=661) Psychosomatic 2.41 (0.64) Biological 2.35 (0.69) PTG (range 1-5) 4.1 (0.8) Optimism (range 3-13) 10.2 (2 .0) Pessimism (range 3-12) 6.1 (2.5) Depression (range 0-45) 11.02(9.70) Note. PTG=posttraumatic growth; ART=antiretroviral therapy. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (SD=273.2) and viral load was 60,520.5 (SD=247,808.6). The average years since HIV diagnosis was 6.4 (SD=4.2). Optimism and pessimism. Optimism and pessimism scores ranged from 3- 12. Participants were likely to endorse the optimism items (mean=10.2, SD=2.0). For example, at baseline, half of the participants scored 11 or 12. Although this is a large number of high scores, the distribution of optimism was not that abnormal (skewness=-l.l4). Pessimism was more normally distributed and had more variance than optimism (mean=6 .1, SD=2.5). Hypothesis 1 . PTG Hypothesis 1, that PTG would occur, was supported. Because PTG responses could range from (1) highly negative change to (5) highly positive change (neutral was scored 3), participants with a mean score of 4 or greater were considered to have experienced PTG. The overall mean PTG score was 4.05 (SD=.77) and 3.96 (SD=.80) for baseline (n=835) and follow-up (n=434), respectively. Fifty-nine percent at baseline and 55% at follow-up scored 4 or higher, indicating that participants, on average, experienced PTG (i.e., moderately positive changes). Supporting the negative/positive response format, 9% o f the sample at both time points had a PTG score less than 3. These participants would have been categorized as “no change” when using a positively valenced response format when in fact they experienced negative changes. Levels of endorsement for each item at baseline ranged from 81% reporting positive changes in their priorities about what is important in life to 48% reporting positive changes in their religious faith (Table 2). 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2. Levels of Posttraumatic Growth Endorsement at Baseline fn=835V Item % Priorities about what is important in life 81.3 Appreciation for the value of my own 78.2 life My own inner strength 75.4 My compassion for others 74.5 Direction for my life 72.4 Involvement in things that interest me 69.6 My understanding of spiritual matters 64.6 My sense of closeness with others 63.8 Handling my difficulties 62.5 Willingness to express my emotion 62.3 My religious faith 48.0 Note. Indicating moderate or highly positive change since HIV diagnosis, score of > 4. Preliminary Hypothesis Testing (2-6Y Bivariate associations between PTG. optimism, pessimism, and study variables Associations between PTG, optimism, pessimism and the other study variables were examined with t-tests for categorical variables with the exception of ethnic differences, which were examined with analyses of variance (Table 3). Associations with continuous variables were examined with Zero-order correlations (Table 4). Demographics. Contrary to expectations, there was a significant inverse correlation between age and PTG such that older participants experienced less PTG, although time since diagnosis was positively associated with age (r=.32, g<.001). As hypothesized, females had significantly higher PTG scores than males. Those with low income (<$15,000) scored significantly higher in pessimism than those with 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3. in PTG. Optimism, and Pessimism at Baseline (n=8351 Variable PTG Optimism Pessimism Demographics Ethnicity1 F(3, 831)= 8,33*** F(3, 831)= 23.58*** F(3, 831)= 4.58** White 3.90a (0.76) 9.57c (2.07) 5.7 la (2.38) Hispanic 4,19b(0.75) 10.81a(1.69) 6.15a b (2.56) African-American 4.13ab(0,76) 10.2 lab( 1.92) 6.60b(2.55) Other 3,97ab(0.74) 9.7 lbc( 1.90) 6.10ab(2.03) Gender t(833)=3.44*** t(833)=1.37 t(833)=0,46 Male 4.01 (0.76) 10.11 (2.00) 6.03 (2.48) Female 4.28 (0.77) 10.39(1.79) 6.15(2.38) Income t(833)=l,56 t(833)=0.87 t(833)=3,85*** <$15,000 4.07 (0.79) 10.19(2.01) 6.27 (2.57) $15,000+ 3.99(0.71) 10,06(1.91) 5,58 (2.18) Medical status On Antiretrovirals t(833)=-3.33*** t(833)=l .96* t(833)=0.28 No 3.87 (0.84) 9.88 (2.13) 6.10(2.42) Yes 4.09 (0,74) 10.22(1.93) 6.04 (2.48) Viral load t(833)=3.26** t(833)=2.34* t(833)=-2.76** Undetectable (<500) 4.14(0.72) 10,33 (1.85) 5.79 (2.39) Detectable (500+) 3.97 (0.79) 10.00 (2.06) 6.26 (2.52) CD4 count t(833)=-1.09 t(833)=-2.01* t(833)=2.58** <200 3,99 (0.80) 9.89 (2.03) 6.46 (2.46) 200+ 4.06 (0.76) 10.22(1.95) 5.93 (2.46) - J Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3 (cont’d) Variable PTG Optimism Pessimism Health behavior Illicit drug use t(833)=4.46*** t(833)=4.75*** t(833)=-2.30* No 4.13(0.75) 10.36(1.87) 5.92 (2.44) Yes 3.87 (0.78) 9.66 (2.11) 6.35 (2.52) Smoker t(833)=2.83** t(833)=3.02** t(833)=-2.61** No 4.12(0.78) 10.34 (2.05) 5.85 (2.60) Yes 3.89 (0.86) 9.92 (2.17) 6.04 (2.69) Variables relevant to those on ART (n=66H 95% adherence t(659)=-2.16* t(659)=-1.61 t(659)-2.34* No 3,87 (0.74) 9.68 (2.07) 6.03 (2.58) Yes 4.06 (0.73) 10.12( 1.88) 5.70 (2.44) Note. PTG=posttraumatic growth; ART=antiretroviral therapy; *p<05, **p <01, ***p< 001; 'Means with different superscripts are significantly different. l/« 00 Table 4. Correlations Between Major Variables fn=835') Variable PTG Optimism Pessimism Demographies Age -.07* -.05 -.07 Medical status Years since HIV diagnosis -.07* -.10** -.02 CD4 count .03 .08* -.04 CD4 count time 2 (n=744) .04 .04 -.05 Viral load (log) -.10** _ 09** .10** Viral load (log) time 2 (n=744) -.09* -.03 . 10** Health Behaviors Cigarette use _ [9*** 1^*** .07* Exercise .07* .07* -.10** Healthy diet .08* -.04 _ 14*** Alcohol use -.18* -.05 -.05 Psvcholoeical variables Depression -.35*** - 37*** .36*** Depression time 2 (n=434) , 21*** . 22*** .27*** Optimism — _ 34*** Optimism time 2 (n=434) 24*** 4 9 *** . 23*** Pessimism . 20*** — — Pessimism time 2 (n=434) . 14** -.13** 42*** PTG time 2 (n=434) 41*** .16** -.12* Amone those on ART fn=6 6 n Total side effects experienced -.03 _ j7*** . 11** Psychosomatic -.06 -.18*** .10* Biological -.00 _ 11** .11** Difficulty dealing with side -.07 -.09* .01 effects Psychosomatic -.07 -.09* -.01 Biological -.06 -.10* .04 Note. PTG=posttraumatic growth; ART=antiretroviral therapy; *p<.05, **p <.01, ***p<001. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with cigarette use and inversely associated with exercising and eating a healthy diet (all a ’s< 05). Psychological variables. PTG and optimism were inversely associated with and pessimism was positively associated with depression (all p ’s< 001). As expected, PTG and optimism were positively associated whereas both PTG and optimism were inversely associated with pessimism. All o f these variables had moderate correlations (.41- 49) between baseline and follow-up scores, indicating stability in these constructs over time. Among those on ART. Among the 661 participant on ART at baseline, PTG was positively associated and pessimism was inversely associated with higher levels of adherence. Optimism was inversely associated with the absolute number of side effects experienced and the difficulty in dealing with side effects. Pessimism was positively associated with the absolute number o f side effects experienced, but was unrelated to the difficulty in dealing with side effects. Multivariable statistical models Hypothesis 2: Disease status and progression Hypothesis 2 indicated that, compared with those scoring high in pessimism or low in optimism and PTG, those scoring low in pessimism or high in optimism and PTG would display greater improvements in their physical health status (or show stable or less disease progression). Two biological indicators o f physical health were examined: CD4 and Viral load counts. The associations between PTG, optimism, pessimism and disease status were examined in a series of multivariable regression 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analyses. First, concurrent relationships were examined between these variables at baseline. Second, predictive relationships were examined for follow-up disease status (CD4/viral load), after adjusting for baseline disease status. If PTG, optimism, or pessimism were significantly associated with disease status, depression and health behaviors were examined in additional multivariable regression models as potential mediators of these relationships. After examining these multivariable models, interaction terms were created by multiplying each variable by PTG, pessimism, and optimism (e.g., PTG by gender) and were added to each model. These analyses determined whether any variable moderated the relationship between PTG, optimism, pessimism and the outcome variables. Changes in PTG were examined in relation to viral load and CD4 counts over time in regression analyses that included PTG subgroups. The PTG subgroups depended on PTG scores at time I and time 2. These analyses used the smaller follow-up sample, which had complete longitudinal data available for PTG (n=428 with complete biological data). Using a cut-point of 4 at time 1 and time 2, 4 PTG groups were formed: those who always experienced PTG (high-high, n=180), those who never experienced PTG (low-low, n=l 17), those who gained PTG (low-high, n=58), and those who lost PTG (high-low, n=79). Finally, a slope analysis was utilized to examine changes in CD4 cell count and viral load over time. Up to 6 data points per participant were possible, up to 3 data points (i.e., most recent, 2n d most recent, 3rd most recent) at baseline and follow- up, respectively. A simple linear regression model was fitted for CD4 and viral load 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. data against calendar time in months (or a fraction thereof) for each participant. This yielded a slope for each participant indicating the rate of change in CD4 and viral load levels over time. Slopes were not calculated for participants with fewer than 3 valid measurements, yielding an analytical subgroup sample o f 730. PTG, optimism, and pessimism were then examined in relation to the rate of change in CD4 counts and viral load (statistically controlling for the same covariates as the previous models including initial CD4/viral load levels). Each of these analyses included clinic status, age, gender, ethnicity, SES, antiretroviral use, CD4 count, viral load, depression, PTG, pessimism, and optimism in the model. ART use was dummy coded with no use as the reference for the baseline analysis and no use at time 1 & 2 as the reference for the follow-up analysis. Fifty-four percent (n=404) were always on ART, 22 percent (n=165) were on ART at baseline and ART status was unknown at follow-up, 6 percent (n=44) were not on ART at baseline and ART status was unknown at follow-up, 7 percent (n=5l) were never on ART, 6 percent (n=41) started ART between baseline and follow-up, and 5 percent (n=39) stopped ART between baseline and follow-up. The main effect models for the baseline and longitudinal analyses are located in Table S. Viral load Baseline analyses. Neither PTG nor optimism was significantly associated with viral load in the baseline multivariable model (Table 5). Although pessimism was significantly associated with viral load at baseline without depression in the model (B=.07, p<_05), once depression was added to the model it became marginally 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significant (p<lO ) suggesting that depression mediated the concurrent relationship between pessimism and viral load. A single significant interaction term at baseline emerged between CD4 count and PTG (g< 05). Among those with low CD4 counts, PTG was inversely associated with viral load. There was no relationship between PTG and viral load among those with high CD4 counts (Figure 2). Longitudinal analyses. On average, participant’s viral load fell by 9,818 (SD = 27,6108) from 59,061 to 49,243 over time. Baseline levels of viral load and the use of antiretrovirals accounted for the majority o f this change. However, pessimism was a significant independent predictor of higher levels of viral load after controlling for all covariates and baseline values of viral load (Table 5). The strength o f this finding did not change with/without depression in the model, indicating that depression had no mediating role in this relationship over time. Adding health behaviors (smoking, alcohol/drug use, exercising, and healthy diet) to this model also did not diminish this finding. Adding interaction terms to this model yielded 2 significant interactions. First, there was an interaction between PTG and pessimism (g<01), indicating that PTG was associated with lower levels o f viral load over time for those scoring low pessimism (Figure 3). That is, among those scoring low in pessimism, those scoring high in PTG had significantly lower levels o f viral load compared to those scoring low in PTG. The second significant interaction was between gender and optimism (£<05), and indicated that gender moderated the relationship between optimism and 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5 Multiple Regression Analyses of Disease Status Viral Load CD4 count Predictor Time I Time 2 Time 1 Time 2 CD4 count T1 _ 22*** _ ]2 *** NA yg*** Viral load T l NA .38*** - 38*** -.02 Age -.12*** .00 -.03 .01 Gender (l=male) Ethnicity (vs. white) .01 .02 -.03 -.01 Hispanic -.04 .04 -.09* -.04 African-American .03 .08* -.04 -.03 Other -.02 .06* -.04 -.03 Income (1=>$15,000) .04 -.05 .04 .05* Years since HIV diagnosis ART use .04 .04 -.05 -.05" On ART T l & T2 - 34*** _ 25*** -.07" . 11* Stopped T2 NA .01 NA .02 Started T2 NA - 29*** NA 11*** On ART T l, T2 Unknown NA -.18** NA .06 No ART T l, T2 Unknown NA -.06 NA .03 Depression .08* -.01 .02 .00 PTG -.01 -.03 -.02 .02 Optimism .03 .04 .08* -.03 Pessimism .06" .07* .02 -.01 R2 .30 .38 .18 .63 D f 18, 816 23, 720 18, 816 23, 720 F 19 92*** 19.07*** 9.98*** 54.05*** Note. PTG=posttraumatic growth; ART=antiretro viral therapy; these are standardized beta’s; NA=Not Applied; These analyses also controlled for clinic status; "p<. 10, *_p<.05, **g< 01, ***p< 001; ART use was dummy coded with no use as the reference for the baseline analysis and no use at time 1 & 2 as the reference for the follow-up analysis. 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2. PTG was associated with lower levels o f viral load at baseline among those with low CD4 counts 12 CD4<200 CD4 200+ Time I Figure 3. PTG had a positive impact on viral load at follow- up among those scoring low in pessimism Low pessimism High pessimism Time 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. viral load (Figure 4). That is, an inverse relationship between optimism and viral load held for females only; there was no relationship between optimism and viral load for males. Changes in PTG. Change in PTG was not associated with viral load over time in an analysis of covariance model that included the same covariates as the prior models. The pessimism by PTG interaction term found in the prior longitudinal model was not significant when it was added to this model. For exploratory purposes, a more liberal cutoff of 3 instead o f 4 was used to determine PTG change groups. This new PTG change variable was not associated with viral load over time. Figure 4. Optimism had a positive impact on viral load at follow-up for females El Low optimism M High optimism Female Male Gender 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Viral load slope. On average, viral load levels dropped between first and last viral load lab results (mean viral load log slope=-0.05, SD=0.53; mean follow-up of=2.0l years, SD=0.59). PTG, optimism, and pessimism were not significantly associated with the slope of viral load in the multivariable model (model F(23, 706)=7.70, p< 001, R2 = 20). None o f the interactions that were significant in the prior longitudinal model of viral load were significant in this model. CD4 count Baseline analyses. Optimism had a significant positive association with CD4 count in the baseline multivariable model, regardless o f whether depression was in the model (Table 5). Adding health behaviors (e.g. exercise, diet, etc.) to this model did not affect this finding. There were two significant interactions at baseline. The first interaction was between gender and pessimism (p< 01), such that females who scored high in pessimism had the highest CD4 counts (Figure 5). Second, there was a significant interaction between depression and pessimism (p<0l). Among those scoring high in depression, those scoring high in pessimism had significantly lower CD4 counts than those scoring low in pessimism (Figure 6). Longitudinal analyses. On average, participant’s CD4 count rose by 30.74 (SD = 181.22) from 405.21 to 435.94 over time. Baseline levels of CD4 and the use o f antiretrovirals accounted for the majority of this change. There were no significant main effects for PTG, optimism, or pessimism in CD4 count over time. 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5. Females scoring high in pessimism had the highest CD4 counts at baseline 460 - ] 440 - o E 420 - 3 400 - 8 'T Q 380 - O 360 - 340 - D Low pessimism ■ High pessimism Female Male Figure 6 . Those high in depression and pessimism had the lowest CD4 counts at baseline 450 U Low pessimism ■ High pessimism Low depression High depression Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, there were significant two-way interactions. First, there was a significant interaction between Hispanic ethnicity and PTG (p< 001). Among Hispanics there was a significant positive association between PTG and CD4 count over time (Figure 7). That is, Hispanics who experienced PTG had a larger increase in CD4 count compared to Hispanics who did not experience PTG. Second, there was a significant interaction between optimism and PTG (p< 01). PTG predicted positive changes in CD4 counts among those with moderate scores in dispositional optimism (Figure 8). That is, those who experienced PTG and had moderate optimism scores had a significantly larger increase in CD4 count compared all other groups. Figure 7. Hispanics who experienced PTG had a significant increase in CD4 counts over time I El Low African- Hispanic American White Ethnicity Other 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8. PTG had a beneficial influence on CD4 count over time among those with moderate levels of optimism C M S 550 s < Low Moderate High optimism optimism optimism Time 1 < D C o u • 'T Q U T 3 « a ? 3 < Figure 9. Those with undetectable viral loads and low pessimism scores had the highest CD4 counts over time CM 400 - pessimism High pessimism Undetectable viral load (<500) Detectable viral load (500+) Time 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Third, there was a significant interaction between pessimism and viral load (p< 001). Among those low in viral load at baseline, pessimism was inversely associated with CD4 over time. Among those high in viral load at baseline, pessimism was unrelated to CD4 over time (Figure 9). Fourth, there was an interaction between baseline CD4 count and pessimism (p<001). Among those with moderate CD4 levels (200-500) at baseline, those scoring high in pessimism had lower CD4 counts at follow-up compared to those scoring low in pessimism. Because this difference in CD4 count was relatively small (<30 points) and did not appear meaningful this interaction is not presented graphically. Changes in PTG. Change in PTG was not associated with CD4 counts over time in an analysis o f covariance model that included the same covariates as the prior models. However, the Hispanic by PTG interaction term was significant when it was added to the model (n=428, mean square=72610.10, df=3, F=2.69, p<.05). Post hoc analysis indicated that Hispanics who always experienced PTG had significantly higher CD4 counts over time compared to Hispanics who never experienced PTG (p=.0 l). The means for these groups are presented graphically in Figure 10. The optimism by PTG interaction was not significant in this model. For exploratory purposes, a more liberal cutoff of 3 instead of 4 was used to determine PTG change groups. In an analysis o f covariance model that included the same covariates as the prior models, this variable was marginally significant in predicting CD4 count over time for the entire sample (n=428, mean square = 62951.67, df=3, F=2.33, p=.07). Posthoc analyses (Figure 11) revealed that those 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 10. Relationship between changes in PTG and CD4 counts over time among Hispanics 550 500 450 a 3 J 400 Q U 350 300 250 Time 2 Time 1 Always experienced PTG(n=14) Lost PTG (n=38) Gained PTG (n=58) • Never experienced PTG(n=33) Figure 11. Relationship between changes in PTG and CD4 counts over time 480 460 440 _ 420 c 3 O O S 380 O 360 340 400 320 300 Time 1 Time 2 -Always experienced PTG(n=362) -Gained PTG (n=25) -Never experienced PTG(n=l5) - Lost PTG (n=26) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. who always experienced PTG had significantly higher CD4 counts over time compared to those who never experienced PTG (p=.06) and those who lost PTG (p=.05). Slope analyses. On average, CD4 counts rose between first and last CD4 count lab results (mean CD4 slope=2.52, SD=16.56; mean follow-up of 2.15 years, SD=0.68). The multivariable CD4 slope model was significant (F(23, 706)=2.91, B< 001, R2 =.09), and indicated that PTG was marginally associated with an increase in CD4 slope (B= 07, t=1.72, p=.08). Furthermore, the interactions that were significant in the prior longitudinal model were also significant in the same directions when added to this model (with the exception of the pessimism by CD4 count and pessimism by viral load interactions). These interactions were between PTG and Hispanic ethnicity (p< 05), and PTG and optimism (g<05). Summary of disease status and progression analyses Hypothesis 2 was partially supported. These results indicate that pessimism has deleterious effects on health; pessimism predicted higher levels of viral load and lower levels of CD4 counts over time in the longitudinal analyses, but not the slope analyses. Neither health behaviors (diet, exercise, smoking, alcohol, or illicit drug use) nor depression mediated these relationships. The inverse relationship with CD4 count over time was strongest among those healthier at baseline (i.e., those with undetectable viral loads). This link with health is consistent with the notion that psychological variables have a greater impact earlier in the disease course (Scheier & Bridges, 1995). The significant interaction between depression and pessimism at 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. baseline suggested that the sickest participants (i.e., those with the lowest CD4 counts) were more likely to be depressed and pessimistic. The relationships between disease progression, optimism, and PTG were more complex. Ethnicity moderated the impact of PTG on CD4 count over time, PTG predicted positive CD4 changes for Hispanics only. That is, Hispanics who experienced PTG experienced a larger increase in CD4 count compared to Hispanics who did not experience PTG. Importantly, this relationship among Hispanics remained significant when examining changes in PTG over time and when examining the slope of CD4 counts over time. Those who experienced PTG at both time points had significantly higher CD4 counts over time compared to those who never experienced PTG. Because Hispanics had the highest PTG scores of all ethnic groups it may be that large amounts of PTG are necessary in order for it to have a protective relationship with the immune system. However, the only main effect for PTG on CD4 count was found in the slope analysis (marginally significant). In order to determine whether acculturation played a role in this relationship, the longitudinal analyses were repeated among Hispanics stratifying by the language of the survey (Spanish vs. English versions). Although the PTG CD4 count relationship was stronger for the Spanish survey, this finding was not consistent across analyses. It is unclear whether there are any specific facilitators of this relationship within the Mexican, Central American, or South American cultures. For example, it may be that when Hispanics experience PTG they are more likely to turn to religion. However, when the 2 religious oriented items (“my understanding of spiritual 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. matters” and “my religious faith”) were removed from the PTG scale, the significant relationship with CD4 count remained. Thus, other co-factors must be involved. Interestingly, changes in PTG over time were associated (although marginally) with CD4 counts for the entire sample when a more liberal PTG cut- point was used to form the groups. These results suggest that underlying process of PTG, experiencing and maintaining positive changes, may be important in relation to disease progression regardless of ethnicity. The combination of experiencing PTG and not anticipating negative outcomes (i.e., low pessimism scores) had a beneficial impact on viral load over time. That is, perceiving positive changes was beneficial when avoiding negative outcome expectancies. This finding suggests that pessimistic thinking inhibits the salutary influence of PTG. This interaction did not hold in the slope analysis or when examining changes in PTG overtime, although the sample sizes were smaller for these analyses. Those who experienced PTG with moderate optimism scores had the largest positive changes in CD4 count. That is, perceiving positive changes in life and holding mild positive outcomes expectancies had a beneficial impact on CD4 counts. This represents current reflections on positive life changes since diagnosis and a mild future orientation towards positive outcomes. This finding suggests that extremely high levels o f positive expectancies can inhibit the salutary influence o f PTG. Indeed, a degree o f realism may be necessary for optimistic thinking to have positive influences on health. It may be that those who strongly expected positive outcomes 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were unable to achieve those outcomes, either through maladaptive behaviors (e.g., denial, avoidance) or a lack of effort (e.g., complacency), resulting in increased levels o f frustration and stress that may have stalled health improvements. Likewise, those who held no positive outcome expectancies may have lacked the motivation or encouragement needed to improve their health, and thus had less health improvement. Those with moderate levels of positive outcome expectancies were likely more realistic about their situations and worked harder at achieving their goals. However, behaviors other than diet, exercise, smoking, illicit drug use, and alcohol use need to be considered because adding these health behavior variables to the statistical equation did not affect this finding. Whether optimism had a curvilinear relationship (i.e. inverted U) with CD4 counts was explored by adding squared optimism scores to the model. This variable was not significant (B=-.25, t=-1.54, g=. 12). Nevertheless, the combination o f experiencing PTG and moderate levels o f optimism was particularly helpful in increasing CD4 counts. It is unclear why optimism had a beneficial influence on viral load at follow- up for females, but not for males. This difference may be important because studies find gender differences in viral load levels and disease progression (e.g., Anastos et al., 2000). It may be that males and females perceive extremely high or low positive outcome expectancies differently. For example, females who strongly expected positive outcomes may have been more likely to achieve those outcomes (e.g., through adaptive behaviors) than males. However, behaviors other than diet, exercise, smoking, illicit drug use, and alcohol use need to be considered because 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. adding these health behavior variables to the statistical equation did not affect this finding. At baseline, among those with severely damaged immune systems (CD4 count <200, an AIDS defining criteria), PTG was associated with less viral load. Thus, those patients who were sicker yet had their viral loads under control perceived more benefits from their condition. Because this interaction was not significant in the follow-up analysis, this difference did not influence disease progression. At baseline, females who had less immune system damage (i.e., highest levels o f CD4 counts) were more likely to be pessimistic. It is unclear why these relatively healthy females anticipated negative fUture outcomes. Because females had significantly higher pessimism scores than males, it may be that females are more likely to hold negative expectancies similar to the findings that females consistently score higher in depression than males (Nolen-Hoeksema, 1987). Although pessimism was highest for females earlier in the disease course, because pessimism did not interact with gender at follow-up this difference did not have a meaningful influence on changes in CD4 over time. In summary, pessimism predicted disease progression, especially for those healthier at baseline. PTG predicted CD4 improvement for Hispanics and those with moderate optimism scores. PTG predicted viral load improvements for those with low pessimism scores and optimism predicted viral load improvements for females. 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hypothesis 3: Depression It was hypothesized that optimism and PTG would be inversely associated with depression and pessimism would be positively associated with depression. However, because psychological distress and PTG can co-occur, the relationship between PTG and depression was hypothesized to be weaker than the relationship between depression and optimism or pessimism. In addition, two-way interactions were explored to determine whether any factor moderates the relationships between depression and optimism, pessimism, or PTG. Baseline analyses. Baseline regression analyses (Table 6) indicated that PTG, optimism, and pessimism were significantly associated with depression in the expected directions. There were two significant interactions at baseline. First, there was an interaction between gender and optimism (p< 0l). Females scoring low in optimism had the highest levels of depression (Figure 12). Second, there was an interaction between viral load and optimism (p< 01). Those with detectable viral loads and low optimism scores had the highest depression scores (Figure 13). Longitudinal analyses. In the follow-up analysis (Table 6), only the pessimism main effect remained a significant predictor of depression. However, there were 2 significant interaction terms. First, the relationship between pessimism and depression at follow-up was moderated by baseline depression (j><05). That is, the positive relationship between pessimism and depression was only among those scoring low in depression at baseline (Figure 14). Second, there was a significant 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6 Multiple Regression Analyses of Depression____________________ Predictor Time 1 Time 2 Depression T1 NA .50*** Age -.04 .06 Gender (l=male) -.08** .00 Ethnicity (vs. white) Hispanic .04 -.01 African-American .07 .01 Other .04 .08* Income -.09** -.09* Years since HIV diagnosis .05 .01 CD4 count .02 -.02 Viral load (log) .08* -.03 ART use On ART -.06 -.14* Stopped T2 NA -.02 Started T2 NA -.05 PTG . 24*** .02 Optimism -.22*** .06 Pessimism 2 j *** .10* R2 .30 .36 Df 18, 816 21,412 F 19.65*** 11.10*** Note. PTG=posttraumatic growth; ART=antiretroviral therapy; T1 = baseline. These are standardized beta’s. Clinic status was also controlled for. T p=.06, *p<05 **p <01, ***p<0 0 l. ART use was dummy coded with no use at time 1 as the reference for the baseline analysis and no use at time 1 & 2 as the reference for the follow-up analysis. 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 12. Females scoring low in optimism had the highest levels of depression at baseline 17 Female Male M Low optimism M High optimism i I I Figure 13. Those with low optimism scores and detectable viral loads had the highest depression levels at baseline 15 H Low optimism ■ High optimism Undetectable viral load Detectable viral load (<500) (500+) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interaction between starting to use ART and optimism (p<05). Those who were optimistic at baseline and started ART between baseline and follow-up had the highest depression scores compared to all other groups (Figure 15). Changes in PTG. The relationship between depression over time and changes in PTG was examined a second way. Using a cut-point o f 4 at time 1 and time 2, 4 PTG groups were formed: those who always experienced PTG (high-high), those who never experienced PTG (low-low), those who gained PTG (low-high), and those who lost PTG (high-low). This variable significantly predicted depression over time in an analysis of covariance that included the same covariates as the prior models (mean square=389.21, df=3, F=6.22, p<00l). Post hoc analysis indicated that those who always experienced PTG and those who gained PTG had significantly lower depression scores over time compared to those who never experienced PTG or lost PTG (all p’s<.01). The means for these groups are presented graphically in Figure 16. Summary o f depression analyses. Hypothesis 3 was partially supported. At baseline, PTG, optimism, and pessimism were significantly associated with depression in the hypothesized directions. At follow-up, both pessimism and changes in PTG predicted depression over time. Pessimism influenced depressive symptoms, particularly among those who were less depressed at baseline. This suggests that despite scoring low in current depression, holding negative expectancies about the future can lead to depressive symptoms over time. 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 14. Those scoring low in depression and pessimism at time 1 had the lowest depression scores at time 2 17 M Low pessimism ■ High pessimism Low depression High depression Time 1 Figure 15. Among those scoring high in optimism at baseline, those who started ART use were more likely to become depressed U Low optimism ■ High optimism Always on Never on Started Stopped ART ART using using Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 16. Relationship between PTG and depression over time c _o '35 c/i < u i— a. < u Q 16 15 14 13 12 11 10 9 8 7 6 Time 1 Tune 2 • Never experienced PTG(n=l 17) ■Gained PTG(n=58) ■Lost PTG(n=79) •Always experienced PTG(n=180) Optimism predicted higher levels o f depression over time among those who started ART between baseline and follow-up. This finding is most likely the result of expectancies that do not match outcomes. That is, because optimists expect positive outcomes and starting ART often occurs when health begins to deteriorate (a negative outcome), this is a case where positive expectations went unfulfilled. This setback led to higher levels of depression. Thus, this interaction documents a negative ramification of holding positive expectations that did not materialize. Females who scored high in depression at baseline scored significantly lower in optimism than those who scored low in depression. Thus, females scoring high in depression are less likely to expect positive outcomes in the future. However, 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. because this interaction did not hold in the longitudinal analysis, this difference did not affect levels of depression over time. Those with low optimism scores and detectable viral loads had high depression scores. This suggests that patients who do not have their viral loads suppressed and are depressed are less likely to expect positive outcomes in the future. However, because this interaction did not hold in the longitudinal analysis, this difference did not affect levels o f depression over time. Although baseline levels o f PTG did not significantly predict depression levels over time, change in PTG was a significant predictor. Importantly, those who experienced PTG at both time I and 2 or gained PTG from time 1 to time 2 were less depressed than those who never experienced or lost PTG. This is consistent with previous research (Davis, Nolen-Hoeksema, & Larson, 1998) and suggests that the process of PTG, achieving and maintaining positive changes, is associated with lower levels of depression over time. Hypothesis 4: Correlates and predictors of PTG. optimism, and pessimism Correlates o f PTG found in prior studies were expected to be associated with PTG in this sample. These relationships were examined concurrently as correlates (at baseline) and in a longitudinal model predicting PTG over time. Levels o f optimism and pessimism were also examined concurrently and in longitudinal models to determine whether there were any common cofactors across these constructs. 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PTG The hypothesized relationships included: H4a) Optimism was hypothesized to be positively associated, while pessimism was expected to be inversely associated with PTG, H4b) Females were hypothesized to experience greater levels of PTG than males, H4c) Time since diagnosis was hypothesized to be positively associated with PTG, H4d) SES was hypothesized to be positively associated with PTG, and H4e) Religiosity was hypothesized to be positively associated with PTG. At baseline, PTG had significant negative associations with age, gender, alcohol use, depression, and pessimism (Table 7). PTG had significant positive associations with Hispanic and African-American ethnicities (vs. White), eating a healthy diet, and optimism. At follow-up, after adjusting for baseline levels, depression was a significant predictor o f lower levels of PTG over time (Table 7). Religiosity was positively associated with PTG. Those with high levels of viral load were slightly more likely to experience PTG at follow-up. Summary of PTG results. Hypothesis 4 was only partially supported. Although H4a was supported at baseline, neither baseline optimism nor pessimism significantly predicted PTG over time. This suggests that, although outcome expectancies interact with PTG in predicting disease outcomes, outcome expectancies are not associated with changes in perceiving benefits over time. As expected, females reported greater PTG than males, supporting H4b. However, this relationship did not hold over time, suggesting that gender differences emerge early in the process of experiencing PTG. 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 7 Multiple Regression Analyses o f PTG Predictor Time I Time 2 Time I PTG NA 24*** Age -.07* -.00 Gender (l=male) -.10** -.04 Income (1=>$ 15,000) -.05 .03 Ethnicity (vs. white) Hispanic 4 - 00 o -.06 African-American .11* .00 Other .03 -.06 Years since HIV diagnosis .04 .02 CD4 count -.01 .05 Viral load (log) -.01 .09" ART use On ART .01 .07 Stopped T2 NA -.02 Started T2 NA .02 Cigarette use .01 -.04 Alcohol use -.01 Exercise .03 .04 Healthy diet .08* .05 Illicit drug use -.03 .08 Religiosity time 2 NA Depression (CES-D) _ 24*** -.14* Optimism 14*** -.01 Pessimism -.08* -.01 R2 Df F .24 23,811 11.46*** .26 27, 406 5.37*** Note. PTG=posttraumatic growth; ART=antiretroviral therapy; T1 = baseline. These are standardized beta’s. Clinic status was also controlled for. ”£<.10, *£<05, **g <.01, ***g<00l. ART use was dummy coded with no use at time 1 as the reference for the baseline analysis and no use at time 1 & 2 as the reference for the follow-up analysis. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H4c was not supported; time since diagnosis was unrelated to PTG in these models. Age was also unrelated to PTG in this model. These results are not entirely consistent with previous that finds time to be important in fostering PTG (Cordova, Cunningham, Carlson, & Andrykowski, 2001; Park et al., 1996), although others do not find this association (Tedeschi & Calhoun, 1996). Furthermore, socioemotional selectivity theory (e.g., Carstensen & Fredrickson, 1998) was not supported because disease status was not a significant predictor of PTG; the perception of time did not play a role in developing PTG. That is, increasing closeness to the end of life, as indicated by disease status, was not associated with higher levels o f growth. H4d was not supported; SES was unrelated to PTG in these models. This is inconsistent with previous research (e.g., Cordova et al., 2001) and suggests that, among HIV populations, PTG does not vary across SES groups. H4e was supported; religiosity was associated with PTG at follow-up. Because religiosity was only measured at follow-up the directionality of this relationship cannot be established. Furthermore, because two items within the PTG scale are spiritually oriented, this relationship may be due to conceptual overlap between these variables. However, when this analysis was repeated with these items removed, religiosity continued to be significantly associated with PTG (B=.l 1, p<.05). Examining the comments by participants about the questionnaires that were documented by the interviewers revealed one quote that may help characterize this relationship. One patient said “I am at peace with myself, God made this happen to 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. me for a reason.” However, this comment may reflect meaning-making as opposed to the benefit finding aspects of PTG. Interestingly, PTG was associated with lower levels of alcohol use and higher levels of eating a healthy diet at baseline, suggesting that PTG may co-occur with important health behaviors. This is consistent with previous research that finds PTG to be positively associated with health behaviors (Siegel & Schrimshaw, 2001; Milam, Ritt-Olson, Unger, 2002). Neither of these health behaviors predicted changes in PTG. Only depression was a significant predictor of PTG over time. Those experiencing depressive symptoms were less likely to perceive benefits or positive changes from their infection. Because changes in PTG were associated with depression over time, the relationship between PTG and depression is reciprocal. This finding is unusual because previous research finds the concurrent relationship between depression/distress and PTG to be small or nonexistent (e.g., Cordova et al., 2001; Leham et al., 1993). Viral load was marginally associated with higher levels of PTG over time, suggesting that a poor disease state may facilitate PTG. However, this relationship is probably not meaningful because the zero-order correlation between the log of viral load at time 1 and PTG at time 2 is not significant (r=-.01, p= 80, n=434) and remains unrelated after partialing out PTG scores at time 1 (r=.03, g=.51, n=434). 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Optimism and pessimism Optimism. At baseline, optimism was significantly associated with Hispanic and African-American ethnicity (vs. white), higher CD4 counts, higher levels of exercise, lower levels o f eating a healthy diet and less illicit drug use, lower levels of depression and pessimism, and higher levels of PTG (Table 8). At follow-up, there were two significant predictors of optimism over time (Table 8). First, depression predicted lower levels of optimism. Second, those who stopped ART use had higher scores in optimism versus those who were never on ART. Religiosity (measured at time 2) was also associated with optimism at follow-up. Pessimism. At baseline, pessimism was significantly associated with Hispanic and African-American ethnicity (vs. white), lower income, higher levels of depression and lower levels of optimism and PTG (Table 8). At follow-up, high CD4 levels predicted lower levels of pessimism (Table 8). Depression predicted higher levels o f pessimism. In addition, optimism predicted higher levels of pessimism. Summary o f optimism and pessimism analyses. Optimism was associated with important health behaviors (exercising, and avoiding illicit drug use) at baseline. It is unclear why optimism was inversely associated with eating a healthy diet at baseline, although this is consistent with the zero order correlation between these variables. Religiosity was associated with optimism at follow-up, however, because religiosity was only measured at follow-up the directionality o f this relationship cannot be established. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 8 Multiple Regression Analyses of Optimism and Pessimism Optimism Pessimism Predictor Time 1 Time 2 Time 1 Time 2 Age .02 .06 -.05 -.02 Gender (l=male) -.02 -.05 -.00 -.02 Income (l=>$ 15,000) -.02 .02 -.08* -,08+ Ethnicity (vs. white) Hispanic 25*** .10" 15*** .03 African-American Ig*** .05 .16*** -.01 Other .04 -.04 .03 .08" Years since HIV diagnosis -.00 .01 -.00 -.01 CD4 count .07* .01 .02 -.12* Viral load .04 -,08+ ,07+ -.06 ART use On ART -.01 .05 .06" -.02 Stopped T2 NA .11* NA .06 Started T2 NA .07 NA .01 Cigarette use .00 .03 .02 .02 Alcohol use .02 .07" -.05 -.03 Exercise .06" .02 -.00 -.00 Healthy diet -.07* -.01 -.06" -.06 Illicit drug use -.06* -.01 .01 .01 Religiosity time 2 NA .12** NA -.01 Depression (CES-D) _ 2?*** -.27*** .22*** 2 i*** PTG ^2 *** -.02 -.08* -.01 Optimism time 1 NA .32*** .29*** .11* Pessimism time I _ 2g*** -.02 NA 3 5 * * * R2 .32 .37 .25 .33 Df 23, 811 27, 406 23, 811 27, 406 F 16.80*** 8.97*** 11.57*** 7.30*** Note. PTG=posttraumatic growth; ART=antiretroviral therapy; Tl = baseline. These are standardized beta’s. Clinic status was also controlled for. "p< 10, *p< 05, **g <.01, ***p<.0 0 l. ART use was dummy coded with no use at time 1 as the reference for the baseline analysis and no use at time 1 & 2 as the reference for the follow-up analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hispanics and African-Americans scored higher in optimism and pessimism than did Whites. It is unclear why Whites would score low on both of these constructs. There is little previous research that examines ethnic differences in optimism/pessimism. However, these ethnic differences did not hold over time. Interestingly, depression predicted lower levels o f optimism and higher levels of pessimism over time. This finding is unusual because optimism did not predict changes in depression over time, suggesting that, in an HIV population, negative affect can negatively impact dispositional optimism over time. Because pessimism did predict changes in depression over time, the association between depression and pessimism is reciprocal. Those healthier at baseline (high CD4 counts) had lower levels of pessimism over time. This suggests that higher levels of health reduce negative outcome expectancies. Those who stopped ART became more optimistic. It is likely that there was ART failure among those who stopped taking these medications. At baseline, those who stopped ART had significantly higher viral loads than those who continued (t(350)=-2.32, p<05), although there was no difference in CD4 counts (t(350)=0.42, P = .67). Aside from the failure ART, it is possible that those who stopped ART were experiencing unbearable levels of side-effects. However, those who stopped ART experienced a similar amount of side-effects as did those who remained on ART (t(350)=-0.95, p= 34)., suggesting that side-effects did not play a role in deciding to stop ART. Because stopping ART did not lead to disease progression as indicated 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. by changes in virai load and CD4 count (Table S), it is possible that those who stopped ART became more confident and thus more optimistic about their future. Although optimism predicted greater levels of pessimism over time, this relationship is most likely spurious. The zero-order correlation between optimism (time 1) and pessimism at time 2 is negative and significant (r=-.13, £><.01, n=434) and is insignificant after partialing out time 1 pessimism scores (r=.02, j>=.73, n=434). Hypotheses 5 and 6 : Adherence and medicine side effect reporting Hypotheses 5 and 6 were examined only among those participants currently using ART without missing data on ART variables (n=661). These variables included ART adherence, side effects experienced, and difficulty in dealing with side effects. Because the larger intervention included an adherence component, these analyses were only performed on baseline data. ART adherence For hypothesis 5, PTG and optimism were expected to be positively associated with adherence while pessimism was expected to be negatively associated with adherence. Seventy-five percent (n=499) of those on ART reported high levels of adherence (>95%). Correlates o f adherence were examined in a multivariable logistic regression analyses (1= >95% adherence). Although PTG and pessimism were associated with adherence in the bivariate analyses, there were no significant main effects for adherence in the multivariable model (Table 9). However, there were a number o f significant interactions. First, there was an 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 9 Multiple Regression Analyses of ART Adherence and Side Effect Reporting Side effects experienced Predictor 95% adherence1 Sum Difficulty dealing with Age 1.04** .05 .06 Gender ( 1-male) 0.50* .00 .00 Ethnicity (vs. white) Hispanic 0.62^ -.02 -.06 African-American 0.78 -.08+ -.06 Other 0.47* .01 -.01 Income (I =>$15,000) 1.30 .01 -.08* Years since HIV diagnosis 0.96+ .06 .08* CD4 1.00 -.07 .01 Viral load (log) 0.96 .01 .05 Adherence NA -.01 -.03 Depression 0.98 .25*** .19*** PTG 1.19 NA NA Optimism 1.01 .04 ,08+ Pessimism 0.97 .03 -.04 X ' 2 ‘ ( T8 ) = 65.94*** F (18, 642) = 7.68*** R2 =. 18 F (18, 595) = 8.32*** R2=.20 Note. Odds ratios. These are standardized betas. These models are only among those on antiretroviral therapy. PTG=posttraumatic growth; ART=antiretroviral therapy; NA = Not Applied. Clinic status was also controlled for. "g< 10, *g<05, * * £ < 01, * * * £ < 001. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 17. Those scoring high in pessimism with detectable viral loads had the lowest levels of adherence m Low pessimism ■ High pessimism Figure 18. Those scoring high in pessimism and depression had the lowest levels of adherence H Low pessimism ■ High pessimism Low depression High depression Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 19. Pessimism had an inverse relationship with adherence for all ethnicities except Hispanic U Low pessimism ■ High pessimism African- Hispanic White Other American Figure 20. There was a positive relationship between optimism and adherence for female 100 - |------------------------------------------------------------------------------------------------------------------------------------------- 95 ■ High optimism Female Male Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interaction between pessimism and viral load (g< 01). Those scoring high in pessimism with detectable viral loads had the lowest levels o f adherence (Figure 17). Second, there was an interaction between depression and pessimism (p< 05). Those scoring high in pessimism and depression had the lowest levels o f adherence (Figure 18). Third, there was an interaction between Hispanic ethnicity and pessimism (p < 01), such that an inverse relationship existed between pessimism and adherence among whites but not among Hispanics (Figure 19). Fourth, there was an interaction (g<05) between gender and optimism (Figure 20), such that there was a positive relationship between optimism and adherence among females but not among males. ART side effects For hypothesis 6, optimism was expected to be inversely associated and pessimism was expected to be positively associated with side effect reporting and difficulty in dealing with side effects. These analyses were performed on both total scores as well as on those side effects determined to be more psychosomatic versus biological in origin. Because there were no hypotheses concerning PTG and side effects, PTG was not included in these analyses. These analyses are located in Table 9. In the multivariable analyses examining the sum o f side effect reporting, depression mediated the relationship between pessimism and side effects. That is, without depression in the model, a high level o f pessimism was significantly associated with higher levels o f side effect reporting (B=.10, £<.05). However, when depression was added to the model, pessimism dropped out of significance and 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. depression was significantly associated with side effect reporting (Table 9). There was one significant interaction, between ethnicity and optimism (p< 05). The inverse relationship between optimism and side effects was stronger among African- Americans versus Whites (Figure 21). None of these results changed in the analyses examining only side effects that were psychosomatic in origin (nausea, diarrhea, insomnia, fatigue, and loss of sex drive) or in the analyses examining side effects that were biological in origin (rash, headache, weight redistribution, tingling limbs, and blood in urine). In the multivariable analyses examining the difficulties in dealing with side effects, neither pessimism nor optimism was associated with reported difficulties. Depression was associated with difficulties, but it did not function as a mediator. Although optimism was marginally associated with difficulties in a positive direction, this relationship is spurious. The zero order correlation with side effect difficulties was negative (r=-.09, p<05) and remained negative when individually partialing out the other variables included in the model (depression, pessimism, CD4 count, etc.). There were no significant interactions. These results did not change in the analyses examining only side effects that were psychosomatic in origin (nausea, diarrhea, insomnia, fatigue, and loss of sex drive) or in the analyses examining side effects that were biological in origin (rash, headache, weight redistribution, tingling limbs, and blood in urine). 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Summary o f ART analyses Hypothesis 5 was partially supported. While PTG was unrelated to adherence in this analysis, optimism was associated with adherence for females. Furthermore, pessimism was associated with adherence under a number of conditions. Those scoring high in pessimism with detectable viral loads had the lowest levels of adherence. These participants do not have their viral loads under control, are not adhering to their drug regimens, and were pessimistic about their future. It may be that pessimistic persons are less likely to adhere to their ART regimens and the higher levels of viral load reflects this lack of adherence or ART failure. In addition, those scoring high in depression and pessimism had the lowest levels o f adherence. That is, those who are currently depressed and expect negative outcomes are less likely to adhere to their drug regimens. The relationship between pessimism and adherence was also moderated by ethnicity, such that an inverse relationship existed between pessimism and adherence among whites but not among Hispanics. It is not clear why there would be ethnic differences in the relationship between pessimism and side effects. Hypothesis 6 was not widely supported. Pessimism was associated with side effect reporting through the relationship that pessimism had with depression. Optimism was associated with lower side effect reporting for African-Americans (vs. Whites). It is not clear why there would be ethnic differences in the relationship between optimism and side effect reporting. Neither pessimism nor optimism was associated with difficulties in dealing with side effects. 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 21. The relationship between optimism and side effect reporting was greatest for African-Americans ■ g 4.5 o optimism optimism j African- Hispanic White Other [ American I ________________________________________________________ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION This study examined the relationships between disease progression and adaptation and PTG, optimism, and pessimism over time among a diverse sample of HIV patients from 6 outpatient clinics located throughout California. Whether PTG, optimism, and pessimism had beneficial relationships with disease status/progression and depression, and whether health behaviors/depression mediated the relationships with disease status, was explored. A summary o f all the regression analyses appears in Table 10. The major hypotheses, which sought to answer whether those who had higher levels o f PTG or optimism and lower levels o f pessimism respond/adapt well to HIV, were partially supported and provided a number o f conclusions. Pessimism was associated with disease progression, depression, and medication adherence Pessimists expect negative experiences in the future, and thus are less inclined to actively pursue their goals. Pessimism in HIV-infected patients should be taken seriously because these negative expectations were self-fulfilling; those scoring high in pessimism were not adapting as well as those scoring low in pessimism. Pessimism predicted poorer disease status (higher viral loads and lower CD4 counts, especially for those healthier at baseline), higher levels o f depression, and lower levels o f medication adherence (primarily among those with high levels of depression). These results are consistent with previous research that finds negative expectancies to be associated with the onset o f symptoms (Reed et al., 1999) and mortality (Reed et al., 1994) among HIV/AIDS patients. In addition, the impact of 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 10 Summary of Multiple Regression Analyses Independent Variable Dependent Variable PTG Optimism Pessimism Viral load (log) time 1 a 0 + b Viral load (log) time 2 C d + Viral load slope 0 0 0 CD4 count time 1 0 + e CD4 count time 2 +f + 8 h CD4 slope +f + 8 0 Depression time 1 - t + Depression time 2 J +k +l ART adherence 0 +m o , e ART side effects experienced NA _P +b ART side effects difficulty NA 0 0 Note. Significant findings are noted with a + for positive and - for negative associations. A 0 indicates a null finding, the relationship was examined but was non-significant. PTG=posttraumatic growth; ART=antiretroviral therapy; NA=not applicable. a Primarily for those with low CD4 counts. b Mediated by depression. c Primarily among those scoring low in pessimism. d Among females only and mediated by ART adherence. c Primarily for those scoring high in depression. f Primarily among Hispanics. g Among those with high PTG scores and moderate optimism scores. h Primarily among those with low viral loads at baseline. 1 Primarily among females and those with detectable viral loads. J Changes in PTG. k Among those who started ART. 1 Primarily among those scoring low in depression at baseline. m Among females only. 0 Primarily among those with detectable viral loads. p Primarily among African-Americans. pessimism found in the present study may be stronger than reported because those who were lost to follow-up in the disease status analyses scored significantly higher in pessimism. There was evidence that disease status influences pessimism. When participants were assessed, they were aware of their CD4/HIV antibody status and high levels of CD4 count predicted lower levels o f pessimism. Although it is possible that the relationship between pessimism and CD4 count is explainable by 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. previous declines in health, this relationship is most likely reciprocal because baseline CD4 counts were statistically controlled for in these analyses. Pessimism is different from optimism Consistent with previous research (Lai, 1994; Marshall, Wortman, Kusulas, Hervig, & Vickers, 1992; Raikkonen et al., 1999; Robinson-Whelen, Kim, MacCallum, & Kiecolt-Glaser, 1997), optimism and pessimism were empirically distinct. That is, for example, scoring low in optimism does not mean one scores high in pessimism. There were unique findings for optimism and pessimism, suggesting that negative and positive expectancies can influence outcomes differently. Indeed, the risks o f pessimism outweighed the protective nature of optimism; pessimism was a more robust predictor of both physical and mental health, suggesting that the benefits of optimism are limited and that negative expectancies can be self-fulfilling. For example, the positive relationship between pessimism and depression over time is consistent with prior research (e.g., Bromberger & Matthews, 1996; Epping-Jordan et al., 1999), as is the lack of association between optimism and depression over time (e.g., Marshall & Lang, 1990; Scheier et al., 1989; Scheier et al., 1994; Study 1). Thus, negative expectancies appear to have a larger detriment on mental and physical health than positive expectancies. Further exploration of optimism and pessimism differences is needed; previous research that failed to analyze these sub-scales or examine interactions in relation to disease progression may have overlooked important relationships. 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PTG influenced disease progression and depression, although the underlying process of PTG is unclear The experiencing o f positive changes since HTV diagnosis was prevalent; fifty-nine percent of the participants experienced moderate to high levels of PTG. This is consistent with previous research among cancer patients that show 60-90% to report perceptions of benefits (e.g., Collins, Taylor, & Skokan, 1990; Petrie, Buick, Weinman, & Booth, 1999; Taylor, Lichtman, & Wood, 1984) and that life threatening diagnoses can lead to positive outcomes. In addition, the positive changes representing PTG were highly interrelated; PTG was supported as a unitary scale. This suggests that positive changes in one area may spur positive change in another. PTG was marginally associated with positive CD4 slopes and was significantly associated with higher CD4 counts over time for Hispanics. These relationships are consistent with the Bower et al. (1998) study, which found discovery of meaning in response to bereavement to be associated with slower declines in CD4 counts. The interaction between PTG and Hispanic ethnicity for CD4 count was not explained by differences in depression, health behaviors, or religious changes. One possible interpretation is that acculturation plays a role in the relationship between PTG and health improvements, although this relationship was found among Hispanics regardless of the survey language (Spanish or English). However, because language is only a tentative proxy of acculturation, this relationship needs further examination. It is not clear how culture may facilitate this 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. relationship. It may be that family and social support factors or trust and belief in the physician/health care provider could explain this relationship. It is unlikely that “yay-saying” accounts for this relationship because CD4 counts were not self-report but abstracted from medical charts and thus not susceptible to any positive bias. The relationship between PTG and disease status was also moderated by outcome expectancies. Among this HIV population, perceiving benefits/positive changes had the greatest impact on disease status when negative expectancies were low and positive expectancies were moderate. Thus, the influence o f PTG appears to vary depending on one’s future orientation. That is, for example, the relationship between PTG and disease status did not exist when participants expected negative experiences in the future. Despite appreciating positive changes in the present, one’s future perceptions can dampen the salutary influence o f PTG. Furthermore, the interaction between PTG and moderate optimism scores suggests that a degree of realism is needed for health benefits to occur (this notion is discussed in the section concerning optimism). These interactions need to be examined among other populations to determine whether these particular relationships are generalizable. When does PTG occur? Because time since diagnosis, age, and disease status were all unrelated to PTG over time in the longitudinal multivariable model, it appears that changes in PTG depend on other factors, and may occur suddenly. Because those recently diagnosed (<3 months) were excluded from participating in this study, the largest changes in PTG may occur closer to diagnosis; the inverse 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. zero-order correlation between PTG and years since diagnosis (r=-.07) supports this view. Consistent with Davis et al. (1998), changes in PTG predicted lower levels of depression over time. That is, the process of obtaining and maintaining PTG reduces depressive symptoms. However, these results also show that depression and PTG have a reciprocal relationship because higher levels o f depression predicted lower levels of PTG over time. Thus, depressive symptoms can impede the development of PTG. Although optimism has shown some promise in promoting PTG (see Affleck & Tennen, 1996 for a brief review), there was the possibility that item overlap between the predictor (optimism) and criterion (PTG) accounted for this relationship (Tennen & Affleck, 1998). Because this study used the revised measure of optimism, which does not contain items concerning benefit finding (e.g., “I’m a believer that ‘every cloud has a silver lining’”), the concurrent relationship between optimism and PTG appears valid. However, because optimism and pessimism did not predict changes in PTG, outcome expectancies did not foster or inhibit PTG over time. This might be limited to HIV populations because other research finds a longitudinal relationship between optimism and PTG (Davis et al., 1998) excluding the benefit finding items. In addition, the lack o f these longitudinal relationships, and statistically adjusting for optimism and pessimism in the other models, provides further evidence that PTG does not simply reflect an underlying optimistic 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disposition and that these dispositions do not account for the relationships between PTG and depression or disease status. Because PTG represents positive changes that occur after diagnosis, the positive relationships found between PTG and health behaviors is sensible and changes in health behaviors could be a logical extension of the PTG construct. That is, in addition to positive changes in relationships and life priorities, positive changes in health behaviors (diet, exercise, etc.) can also stem from diagnosis (e.g., Collins et al., 2001; Siegel & Schrimshaw, 2001). Most importantly in relation to HIV, PTG was associated with ART use and adherence to ART in bivariate analyses. PTG may be associated with health behaviors through supportive others and clearer life goals. However, these relationships with health behaviors, did not entirely account for the relationships between PTG and disease status. Religiosity was associated with higher levels of PTG and optimism. One’s religious faith can provide a theodicy through which one can view their trauma with a positive outlook. Religious beliefs can aid in the adaptation to disease through enacting positive changes, improving outlook in life, and discovering personal strengths and meaning during difficult times (e.g., Dull & Skokan, 1995). Religiosity can therefore predispose one towards greater or less PTG or optimism by providing a framework through which traumas can be appraised. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The relationships between optimism and outcome variables were influenced by other variables The findings for optimism and disease progression suggest a divergent relationship with the immune system. This study suggests that those who score at the extreme end o f the optimism scale are less likely improve their health than those with moderate scores. Previous research provides some clues as to why this might be. For example, there is evidence that when stress is chronic optimists show lower levels o f immunity (Cohen et al., 1999). That is, optimists may have difficulties in situations of prolonged difficulties that are inconsistent with their expectations. This is also consistent with the findings from a study in which optimists showed a reduction in immunoreactivity (NK activity) subsequent to uncontrollable stress (Sieber et al., 1992). Because HIV is a chronic disease with a plethora of disease related stressors (treatment decisions, disclosure, stigma, side effects, etc.), many of which are uncontrollable, it may be that expectations of the extreme optimist exceed reality. The interaction between optimism and ART use for depression supports this view; among those who started ART between baseline and follow-up, those scoring high in optimism had higher depression scores over time compared to those scoring low in optimism. Because optimists expect positive outcomes, and starting ART occurs when health deteriorates (a negative outcome), depressive symptoms are one consequence o f positive expectations that are not fulfilled. This is consistent with research that indicates that unfulfilled optimistic beliefs heighten negative affect (Diener et al., 1991). 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Research investigating mindsets also support the notion that high levels o f optimism may be illusory. When individuals make important decisions they enter a deliberative mindset (i.e., to deliberate over a course o f action). Furthermore, levels of unrealistic optimism are suspended or reduced when individuals enter this mindset (Talyor & Gollwitzer, 1995). That is, people are more realistic when making decisions. Once these decisions are made, however, people enter implemental mindsets where unrealistic optimism returns and this illusory thinking helps foster goal attainment. Thus, a window of realism opens when people enter deliberative mindsets in order to make important decisions. It may be that this window of realism generalizes to global measures of optimism and that high levels of optimism are reduced when important decisions are made. Stopping/starting ART are major treatment decisions that require meaningful deliberation. Those who stopped/started ART between baseline and follow-up were likely to be deliberating this decision during the baseline interview. In addition, those who stopped/started ART had lower baseline optimism scores and higher adjusted follow-up scores than those who were always or never on ART and this change in optimism over time was significant for those who stopped ART. This change in optimism may reflect the underlying decision making process, that is, a reduction in optimism during the deliberative mindset and a subsequent increase in optimism once this decision has been implemented. An alternative explanation is that those participants who stopped ART may have became more optimistic because, although they may have expected health 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. declines subsequent to stopping ART, there were no significant changes in disease status after stopping ART. Thus, optimism may increase following unexpected positive outcomes. However, because it is unknown at what time these participants stopped ART it is unclear whether there were any meaningful changes in health status following ART termination. Likewise, those stopping ART may have felt more control over their situation and thus became more optimistic. Optimism had a beneficial influence on viral load over time for females compared to males. Because this same interaction occurred in the adherence analyses, the possibility that higher levels of adherence accounted for this relationship was examined in a series of regression analyses predicting viral load over time (one model with and one without adherence in the model). This analysis was performed only on those who were on ART at baseline with complete data (n=623) and contained the same covariates as the prior models. Without adherence in the model, the gender by optimism interaction was significant (B=.44, t=1.91, P=.05). When adherence was added to the model (B=-.08, t=-2.36, p= 02), the gender by optimism interaction was reduced and dropped out of significance (B=.39, t=1.71, p=.09) indicating that adherence mediated the relationship between optimism and viral load for females. Thus, for females, optimism is associated with higher levels o f adherence, which in turn is associated with lower levels of viral load over time. Because optimists expect positive experiences in the future, they are more likely to work at attaining their goals. If health improvement is a goal, then ART 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. adherence is an effective route used by optimists for achieving this goal. Importantly, aside from lower levels o f eating healthy, there was little evidence that optimism elicits poor health behaviors. Optimism was associated with overall ART use, exercising, and less cigarette and illicit drug use in the bivariate analyses suggesting that those scoring high in optimism were more likely to utilize behaviors that enable the goal achievement of better health. Depression plaved a role in PTG. optimism/pessimism Depressive symptoms predicted lower levels o f optimism and PTG, and higher levels o f pessimism over time. Thus, higher levels of prior depressive symptoms impede holding positive expectancies and predict greater negative expectancies. This suggests that, among HTV patients, depressive symptoms can negatively change dispositions over time. Because pessimism and changes in PTG predicted depression over time, these particular relationships were reciprocal. It is unusual that a “state-like” measure would predict changes in dispositions, unless those scoring high in depression at baseline were suffering from chronic depression. This is a strong possibility because depression was more stable over time (time 1/time 2 r=.55) than either optimism (timel/time 2 r=.49) or pessimism (timel/time 2 r=.42). There may also be a temporal phenomenon operating, similar that posited by socioemotional selectivity theory (e.g., Carstensen & Fredrickson, 1998), which may explain the relative instability among optimism and pessimism. When people are diagnosed with a life-threatening illness, the future becomes unstable. That is, HIV 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. patients, given their vulnerable health status, may have shorter future horizons. In this case, dispositional optimism may be tapping into expectancies that are not that far in the future (e.g., anticipating positive outcomes next week, as opposed to next year). Thus, both optimism and pessimism may reflect state-like constructs among the terminally ill. Research examining the relationships between dispositional optimism/pessimism and time perspective is needed among these populations. Importantly, because all o f the other analyses controlled for baseline levels of depression, higher levels o f negative affect/depressive symptoms did not account for the disease status results. That is, depressive symptoms and PTG and negative/positive expectancies are empirically distinct. Strengths of the study This study was effective due to its longitudinal design and large, diverse sample. In addition, all of the analyses controlled for a myriad o f important covariates, including ART use and prior CD4 counts and viral loads. These results suggest that perceiving benefits from a life-threatening diagnosis and avoiding negative expectations about the future can have salutary effects on health. Neither health behaviors nor depression mediated o f any o f the longitudinal relationships with disease progression (with the exception of adherence for optimistic females), suggesting that other pathways need to be investigated. Although the relationships between disease progression and PTG, pessimism, and optimism were relatively small, there were a number o f factors working against finding these relationships. Notably, antiretroviral medications (i.e., protease 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. inhibitors) used in conjunction with other antiretrovirals have succeeded in suppressing viral loads to undetectable levels while simultaneously increasing or stabilizing CD4 levels for many patients. Furthermore, given these treatment developments, the follow-up period of this study may not have been be adequate to observe meaningful changes in disease progression. CD4 counts and viral loads improved over time for most participants. In addition, this was a heterogeneous sample, ethnically diverse, with a wide range in SES, and recruited from multiple sites. However, to find significant psychological influences o f disease status despite these hurdles and after adjusting for ART use (which was positively associated with optimism and PTG) and baseline levels o f disease status (i.e., autoregressive effects; time I & 2 CD4 counts r=.78) is particularly impressive and bolsters the confidence in these observed associations. Because a large amount o f variation in HIV progression remains unknown, any additional explanation is important. Indeed, PTG and pessimism warrant further examination, especially with larger sample sizes that are more likely to detect these relationships. Previous research examining the relationships between psychological factors and HIV disease progression indicators has been among homogenous White, affluent men with good access to medical care. However, significant associations were found in this study among a more diverse sample. Although certain relationships may depend on ethnicity, these relationships did not vary across other important social factors (e.g., there were no interactions with SES). These results indicate that, for example, PTG can occur in diverse populations. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Preliminary evidence shows social support satisfaction to be related to PTG (Park et al., 1996). The most likely initial function of social support is as a stress buffer (e.g., Cohen & Wills, 1985), however as the process of PTG unfolds, family and friends may play more active and complex rather than protective roles. Supportive others can teach and nurture traumatized individuals. Because o f the socially interactive nature of PTG, the intervention potential is great. In certain cases, social support may be useful only for interpersonal types of PTG, and not for other types (e.g., change in life priorities). Optimism/pessimism and PTG represent viable factors for future intervention focus. The assumption that optimism and pessimism are traits should not prevent attempts to facilitate changes. Indeed, recent evidence supports the notion that optimism/pessimism can be effective components o f interventions designed to alleviate psychosocial morbidity through reducing distress and enhancing coping and other health behaviors (e.g., Antoni et al., 2001; Mann, 2001). These interventions can lessen the psychological, social, physical, and financial costs of being a patient or a family member thereof. Current and future studies examining health promotion or illness-recovery intervention effects would benefit from including these constructs in their patient assessments. An alternative strategy would be to selectively target pessimists or those scoring low in PTG with psychosocial interventions. For example, among breast cancer patients, there is evidence that a psychosocial intervention can have a greater beneficial effect on less optimistic women (Carver, 1999). Behavior change may be 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dependent on the individual perceiving life and the future as positive enough to be worthy of enacting change. Pessimists have shown to react more favorably towards explicit guidance to engage in effective coping/self-regulation (Cameron & Nicholls, 1998). These studies confirm that pessimists are in greater need of interventions that support the transition from dwelling on an experience to acting on it (i.e., using specific coping strategies). This could entail exercises that encourage an optimistic (and realistic) mindset or focusing on aspects of PTG. Thus, the effective techniques that optimists use to cope may be effectively taught and can change subsequent levels o f optimism and PTG (e.g., Antoni et al., 2001; Calhoun & Tedeschi, 1999; Mann, 2001; Seligman, 1991). One caveat is that positive thinking, or avoiding negative expectancies, is not a panacea. That is, any intervention designed to enhance levels of expectations or perceptions o f positive changes should continue to emphasize traditional health care. Psychological interventions would be more effective when they are integrated within normal health care arenas (e.g., messages delivered by health care providers, seminars given at the clinic before/after medical appointments). In addition, patients should not be chastised for holding negative attitudes. Limitations Although longitudinal, this study was retrospective with regards to PTG. Thus, it could not accurately disentangle the amount of PTG attributed to HIV diagnosis versus the amount due to other unknown factors or of the amount present before diagnosis (although instructions explicitly asked for participants to indicate 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for the time period since diagnosis). Because the PTG scale was close-ended, there may have been other dimensions of positive change not captured by this measure. In addition, with the exception of the biological variables, because measures were in the form of self-report, there is the possibility of socially desirable responding and recall error. There may be alternative explanations for PTG. For example, PTG may represent an emotion-focused coping strategy that allows for positive reappraisal o f one’s situation. Without corroborating reports from family/friends it is unclear whether these changes actually occurred, although they may not be readily observable. However, regardless of whether PTG represents cognitive or observable changes, they remained adaptive. Because a primary outcome of the larger study was sex behavior, sexually inactive patients were excluded from this study. Generalizing these results to sexually inactive HIV patients or other populations should be done cautiously. A large number of participants were not included in the analyses of psychological variables over time, due to being lost to follow-up or responding to a briefer survey that did not include the psychological items. Because those not included in the longitudinal analyses were more likely to have higher viral loads, these results may not generalize to sicker patients. A number of important factors relevant to the processes examined in this study were not assessed. These include coping strategies, various psychological factors (e.g., unrealistic optimism, perceived risk, neuroticism etc.), stress levels, and 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. quality of life. Additional treatments (counseling, anti-depressants, alternative therapies, etc.) undertaken by participants were also not assessed. Health behaviors (e.g., diet, exercise) were assessed with single items and may not have yielded reliable measurements. This study did not adjust the p value despite the many statistical tests performed (i.e. bonferonni). Although certain results may have been produced by chance, significant relationships were in the expected directions and were theoretically meaningful. Conclusion Although a life-threatening diagnosis can lead to many negative outcomes, there are positive factors and strengths that buffer against mental and physical illness. Optimism/pessimism and posttraumatic growth/realizing benefits are two of these protective factors. This study finds evidence o f growth and perceived benefits among a population of HIV/AIDS patients. In addition, experiencing higher levels of growth and lower levels of pessimism/moderate levels of optimism was beneficial in the adaptation to HIV/AIDS over time. Because these factors have a relationship with disease course and important health behaviors, there is the possibility of developing interventions based on these data. This information will not only aid future research examining these processes among similar populations, but will also enable health care professionals to better understand the impact of PTG/optimism on the adjustment process. 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES Adler, N., & Matthews, K. (1994). Health psychology: Why do some people get sick and some stay well? Annual Review o f Psychology. 45. 229-259. Affleck, G., Tennen, H. (1996). Construing benefits from adversity: Adaptational significance and dispositional underpinnings. Journal o f Personality. 64. 899-922. Affleck, G., Tennen, H., Croog S., & Levine, S. (1987). Causal attribution, perceived benefits, and morbidity after a heart attack: An 8-year study. Journal of Consulting and Clinical Psychology. 55. 29-35. Aldwin, C. M., Levenson, M. R., & Spiro, A. (1994). Vulnerability and resilience to combat exposure: can stress have lifelong effects? Psychology and Aging. 9. 34-44. Anastos, K. Gange, S. J., Lau, B., Weiser, B„ Detels, R., Giorgi, J. V., Margolick, J. B., Cohen, M., Phair, J., Melnick, S., Rinaldo, C. R., Kovacks, A., Levine, A., Landesman, S., Young, M., Munoz, Alvaro, & Greenblatt, R. M. (2000). Association of race and gender with HIV-1 RNA levels and immunologic progression. Journal of Acquired Immune Deficiency Syndrome. 24. 218-226. Anderson, E. H. (2000). Self-esteem and optimism in men and women infected with HIV. Nursing Research. 49f5L 262-271. Andersson, G. (1996). The benefits o f optimism: A meta-analytic review of the Life Orientation Tests. Personality and Individual Differences. 21. 719-725. Andersson, G. (1999). Anxiety, optimism and symptom reporting following surgery for acoustic neuroma. Journal of Psychosomatic Research. 46. 257-260. Aspinwall, L. G., & Brunhart, S. M. (1996). Distinguishing optimism from denial: Optimistic beliefs predict attention to health threats. Personality and Social Psychology Bulletin. 10. 993-1003. Aspinwall, L. G., & Taylor, S. E. (1992). Modeling cognitive adaptation: A longitudinal investigation o f the impact of individual differences and coping on college adjustment and performance. Journal o f Personality and Social Psychology. 63, 989-1003. Antoni, M. H., Lehman, J. M., Kilboum, K. M., Boyers, A. E., Culver, J. L., Alferi, S. M., Yount, S. E., McGregor, B. A., Arena, P. L., Harris, S. D., Price, A. A., and Carver, C. S. (2001). Cognitive-behavioral stress management intervention decreases the prevalence o f depression and enhances benefit finding among women under treatment for early-stage breast cancer. Health Psychology. 20, 20-32. 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Birenbaum, L.K., & Robinson, M.A. (1991). Family relationships in two types of terminal care. Social Science and Medicine. 32. 95-102. Billingsley, K. D., Waehler, C. A., & Hardin, S. I. (1993). Stability o f optimism and choice of coping strategy. Perceptual and Motor Skills. 76. 91-97. Bjorck, J. P., Hopp, D. P., & Jones, L. W. (1999). Prostate cancer and emotional functioning: Effects of mental adjustment, optimism, and appraisal. Journal of Psychosocial Oncology. 17. 71-85. Bower, J. E., Kemeny, M. E., Taylor, S. E., & Fahey, J. L. (1998). Cognitive processing, discovery of meaning, CD4 decline, and AIDS-related mortality among bereaved HIV-seropositive men. Journal of Consulting and Clinical Psychology. 66. 979-986. Brenes, G. A., Rapp, S. R., & Miller, M. R. (1999, August). Do optimism & pessimism predict physical functioning? Poster session presented at the annual meeting of the American Psychological Association, Boston, MA. Bromberger, J. T., & Matthews, K. A. (1996). A longitudinal study of the effects of pessimism, trait anxiety, and life stress on depressive symptoms in middle-aged women. Psychology and Aging. 11. 207-213. Burack, J.H., Barrett, D.C., Stall, R.D., Chesney, M.A., Ekstrand, M.L., & Coates, T.J. (1993). Depressive symptoms and CD4 lymphocyte decline among HIV- infected men. JAMA. 270. 2568-2573. Byrnes, D. M., Antoni, M. H., Goodkin, K., Efantis-Potter, J., Asthana, D., Simon, T., Munajj, J., Ironson, G., & Fletcher, M. A. (1998). Stressful events, pessimism, natural killer cell cytotoxicity, and cytotoxic/suppressor T cells in HIV+ black women at risk for cervical cancer. Psychosomatic Medicine. 60. 714-721. Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist. 54. 165-181. Carstensen, L. L., & Fredrickson, B. L. (1998). Influence o f HIV status and age on cognitive representations of others. Health Psychology. 17. 494-503. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Carvajal, S. C., Clair, S. D., Nash, S. G., & Evans, R. I. (1998). Relating optimism, hope, and self-esteem to social influences in deterring substance use in adolescents. Journal of Social and Clinical Psychology. 17. 443-465. Carvajal, S. C., Evans, R. I., & Nash, S. G. (2002). Global positive expectancies of the self and adolescents’ substance use avoidance: Testing a social influence mediational model. Manuscript submitted for publication. Carver, C. S., Pozo-Kaderman, C., Harris, S. D., Noriega, V., Scheier, M. F., Robinson, D. S., Ketcham, A. S., Moffat, F. L., & Clark, K. C. (1993). How coping mediates the effect of optimism on distress: A study of women with early stage breast cancer. Journal of Personality and Social Psychology. 65. 375-390. Carver, C. S., Pozo-Kaderman, C., Harris, S. D., Noriega, V., Scheier, M. F., Robinson, D. S., Ketcham, A. S., Moffat, F. L., & Clark, K. C. (1994). Optimism versus pessimism predicts the quality of women’s adjustment to early stage breast cancer. Cancer. 73. 1213-1220. Carver, C. S., & Scheier, M. F. (1981). Attention and Self Regulation: AControl- Theorv Approach to Human Behavior. New York: Springer-Verlag. Chang, E. C. (1996). Cultural differences in optimism, pessimism, and coping: Predictors of subsequent adjustment in Asian American and Caucasian American college students. Journal o f Counseling Psychology. 43. 113-123. Chang, E. C., & Bridewell, W. B. (1998). Irrational beliefs, optimism, pessimism, and psychological distress: A preliminary examination of differential effects in a college population. Journal of Clinical Psychology. 54. 137-142. Charles, K., Sellick, S. M., Montesanto, B., & Mohide, E. A. (1996). Priorities of cancer survivors regarding psychosocial needs. Journal of Psychosocial Oncology. 14, 57-72. Chesney, M.A. & Folkman, S. (1994). Psychological impact of HIV disease and implications for intervention. Psychiatric Clinics of North America. 17. 163-178. Chesney, M., Folkman, S., & Chambers, D. (1996). Coping effectiveness training for men living with HTV: Preliminary findings. International Journal of STD & AIDS. 7. 75-82. Cohen, S., Tyrrell, D.A., & Smith, A.P. (1991). Psychological stress in humans and susceptibility to the common cold. New England Journal of Medicine. 325. 606-612. 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cohen, F., Kearney, K. A., Zegans, L. S., Kemeny, M. E., Neuhaus, J. M., & Stites, D. P. (1999). Differential immune system changes with acute and persistent stress for optimists vs pessimists. Brain. Behavior, and Immunity. 13. 155-174. Cohen, S., & Wills, T. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin. 98. 310-357. Collins, R. L., Kanouse, D. E., Gifford, A. L., Senterfitt, J. W., Schuster, M. A., McCaffrey, D. F., Shapiro, M. F., & Wenger, N. S. (2001). Changes in health- promoting behavior following diagnosis with HIV: Prevalence and correlates in a national probability sample. Health Psychology. 20. 351-360. Collins, R. L., Taylor, S. E., & Skokan, L. A. (1990). A better world or a shattered vision? Changes in life perspectives following victimization. Social Cognition. 8. 263-285. Cordova, M. J., Cunningham, L. L. C., Carlson, C. R., & Andrykowski, M. A. (2001). Posttraumatic growth following breast cancer: A controlled comparison study. Health Psychology. 20. 176-185. Curbow, B., Somerfield, M. R., Baker, F., Wingard, J. R., & Legro, M. W. (1993). Personal changes, dispositional optimism, and psychological adjustment to bone marrow transplantation. Journal of Behavioral Medicine. 16. 423-443. Davis, C. G., Nolen-Hoeksema, S., & Larson, J. (1998). Making sense o f loss and benefiting from the experience: Two construals of meaning. Journal o f Personality and Social Psychology. 75. 561-574. Diener, E., Colvin, C.R., Pavot, W.G., & Allman, A. (1991). The psychic costs of intense positive affect. Journal of Personality and Social Psychology. 61. 492-503. DiMatteo, M.R., Sherboume, C.D., Hays, R.D., Ordway, L., Kravitz, R.L., McGlynn, E.A., Kaplan, S., & Rogers, W.H. (1993). Physicians’ characteristics influence patients’ adherence to medical treatment: Results from the medical outcomes study. Health Psychology. 12. 93-102. Dull, V.T., & Skokan, L.A. (1995). A cognitive model of religion’s influence on health. Journal of Social Issues. 51. 49-64. Epel, E. S., McEwen, B. S., & Ickovics, J. R. (1998). Embodying psychological thriving: Physical thriving in response to stress. Journal o f Social Issues. 54. 301- 322. 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Epping-Jordan, J. E., Compas, B. E., Osowiecki, D. M., Oppedisano, G., Gerhardt, C., Primo, K., & Krag, D. N. (1999). Psychological adjustment in breast cancer: Processes o f emotional distress. Health Psychology. 18. 315-326. Farran, C. J. (1997). Theoretical perspectives concerning positive aspects of caring for elderly persons with dementia: Stress/adaptation and existentialism. The Gerontologist. 37. 250-256. Friedman, L. C., Nelson, D. V., Baer, P. E., Lane, M. Smith, F. E., & Dworkin, R. J. (1992). The relationship of dispositional optimism, daily life stress, and domestic environment to coping methods used by cancer patients. Journal o f Behavioral Medicine. 15. 127-141. Friedman, L. C., Webb, J. A., Bruce, S. Weinberg, A. D., & Cooper, P. H. (1995). Skin cancer prevention and early detection intentions and behavior. American Journal of Preventive Medicine. 11. 59-65. Fry, P. S. (1995). Perfectionism, humor, and optimism as moderators o f health outcomes and determinants of coping styles of women executives. Genetic. Social, and General Psychology Monographs. 121. 211-245. Glaser, R., Kiecolt-Glaser, J.K., Bonneau, R. Malarkey, W., & Hughes, J. (1992). Stress-induced modulation of the immune response to recombinant hepatitis B vaccine. Psychosomatic Medicine. 54. 22-29. Goodkin, K., Blaney, N. T., Feaster, D., Fletcher, M. A., Baum, M. K., Mantero- Atienza, E., Klimas, N. G., Millon, C., Szapocznik, J., & Eisdorfer, C. (1992). Active coping style is associated with natural killer cell cytotoxicity in asymptomatic HIV-1 seropositive homosexual men. Journal of Psychosomatic Research. 36. 635- 650. Gritz, E.R., Wellisch, D.K., Siau, J., & Wang, H. (1990). Long-term effects of testicular cancer on marital relationships. Psvchosomatics. 31. 301-312. Helgeson, V. S. (1999). Applicability of cognitive adaptation theory to predicting adjustment to heart disease after coronary angioplasty. Health Psychology. 18. 561- 569. Helgeson, V. S., & Fritz, H. L. (1999). Cognitive adaptation as a predictor of new coronary events after percutaneous transluminal coronary angioplasty. Psychosomatic Medicine. 61. 488-495. 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hull, J. G., & Mendolia, M. (1991). Modeling the relations of attributional style, expectancies, and depression. Journal o f Personality and Social Psychology. 61. 85- 97. Ickovics, J. R., Hamburger, M. E., Vlahov, D., Schoenbaum, E. E., Schuman, P., Boland, R. J., & Moore, J. (2001). Mortality, CD4 cell count decline, and depressive symptoms among HIV-seropositive women. JAMA. 285. 1466-1474. Janoff-Bulman, R. (1992). Shattered Assumptions: Towards a New Psychology of Trauma. New York, New York: Free Press. Johnson, J. E. (1996). Coping with radiation therapy: Optimism and the effect of preparatory intementions. Research in Nursing and Health. 18. 3-12. Joseph,, S., Williams, R., & Yule, W. (1993). Changes in outlook following disaster: The preliminary development of a measure to assess positive and negative responses. Journal of Traumatic Stress. 6. 271-279. Kang, E., Bjorck, J. P., & Watts, M. S. (1997, March). Illusory optimism: an adaptive or maladaptive mechanism for coping with Acquired Immunodeficiency Syndrome? Poster presented at the California Psychological Association 51st Annual Convention, Pasadena, CA. Kamen-Siegel, L., Rodin, J., Seligman, M. E., & Dwyer, J. (1991). Explanatory style and cell-mediated immunity in elderly men and women. Health Psychology. 10, 229-235. Kiecolt-Glaser, J.K. & Glaser, R. (1992). Psychoneuroimmunology: Can psychological interventions modulate immunity? Journal of Consulting and Clinical Psychology. 60. 569-575. Lai, J. C. (1994). Differential predictive power of the positively versus the negatively worded items o f the Life Orientation Test. Psychological Reports. 75. 1507-1515. Lai, J. C. (1995). The moderating effect of optimism on the relation between hassles and somatic complaints. Psychological Reports. 76. 883-894. Lauver, D. (1994). Care-seeking behavior with breast cancer symptoms in Caucasian and African-American women. Research in Nursing and Health. 17. 421- 431. 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lauver, D., & Rubin, M. (1990). Message framing, dispositional optimism, and follow-up for abnormal Papanicolaou tests. Research in Nursing and Health. 13. 199-207 Lauver, D., & Tak, Y. (1995). Optimism and coping with a breast cancer symptom. Nursing Research. 44. 202-207. Leham, D. R., Davis, C. G., Delongis, A., Wortman, C. B., Bluck, S., Mandel, D. E., & Ellard, J. H. (1993). Positive and negative life changes following bereavement and their relations to adjustment. Journal of Social and Clinical Psychology. 12. 90- 112. Leserman, J., Jackson, E. D., Petitto, J. M., Golden, R. N., Silva, S. G., Perkins, D. O., Cai, J., Folds, J. D., & Evans, D. L. (1999). Progression to AIDS: The effects of stress, depressive symptoms, and social support. Psychosomatic Medicine. 61. 397- 406. Litt, M. D., Tennen, H„ Affleck, G., & Klock, S. (1992). Coping and cognitive factors in adaptation to in vitro fertilization failure. Journal of Behavioral Medicine. 15, 171-187. Long, B. C., & Sangster, J. I. (1993). Dispositional optimism/pessimism and coping strategies: Predictors of psychosocial adjustment o f rheumatoid and osteoarthritis patients. Journal of Applied Social Psychology. 23. 1069-1091. Lyketsos, C.G., Hoover, D R., Guccione, M., Senterfitt, W., Dew, M.A., Wesch, J., VanRaden, M.J., Treisman, G.J., & Morgenstem, H. (1993). Depressive symptoms as predictors of medical outcomes in HIV infection. JAMA. 270. 2563-2567. Magaletta, P. R., & Oliver, J. M. (1999). The hope construct, will, and ways: Their relations with self-efficacy, optimism, and general well-being. Journal of Clinical Psychology. 55. 539-551. Marshall, G. N., & Lang, E. L. (1990). Optimism, self-mastery, and symptoms of depression in women professionals. Journal of Personality and Social Psychology. 59, 132-139. Marshall, G. N., Wortman, C. B., Kusulas, J. W., Hervig, L. K., & Vickers, R. R. (1992). Distinguishing optimism from pessimism: Relations to fundamental dimensions of mood and personality. Journal of Personality and Social Psychology. 62, 1067-1074. 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Maruta, T. M., Colligan, R. C., Mallinchoc, M. & OfFord, K. P. (2000). Optimists vs pessimists: Survival rates among medical patients over a 30-year period. Mavo Clinic Proceedings. 75. 140-143. Milam, J. E., Ritt-Olson, A., & Unger, J. (2002). Posttraumatic Growth among Adolescents. Manuscript under review. Mroczek, D. K., Spiro, A., Aldwin, C. M., Ozer, D. J., & Bosse, R. (1993). Construct validation of optimism and pessimism in older men: Findings from the normative aging study. Health Psychology. 12. 406-409. Mumford, D. M. (1997). Thank God I have cancer. JAMA. 278. 956. Namir, S., Wolcott, D. L., Fawzy, F., & Alumbaugh, M. (1987). Coping with AIDS: Psychological and health implications. Journal of Applied Social Psychology. 17. 309-328. Nolen-Hoeksema, S. (1987). Sex differences in unipolar depression evidence and theory. Psychological Bulletin. 101. 259-282. Nolen-Hoeksema, S., Girgus, J. S., & Seligman, M. E. P. (1992). Predictors and consequences of childhood depressive symptoms: A 5-year longitudinal study. Journal of Abnormal Psychology. 101. 405-422. O ’Brien, W. H, VanEgeren, L., & Mumby, P. B. (1995). Predicting health behaviors using measures of optimism and perceived risk. Health Values. 19. 21-28. O’Leary, V. E., & Ickovics, J. R. (1995). Resilience and thriving in response to challenge: An opportunity for a paradigm shift in women’s health. Women’s Health: Research on Gender. Behavior, and Policy. 1. 121-142. Park, C. L., Cohen, L. H., & Murch, R. L. (1996). Assessment and prediction of stress-related growth. Journal o f Personality. 64. 72-105. Paterson, D., Swindells, S., Mohr, J., Brester, M., Vergis, R., Squier, C., Wagener, M., & Singh, M. (2000). Adherence to protease inhibitor therapy and outcomes in patients with HTV infection. Annals o f Internal Medicine. 133. 21-30. Peterson, C., Seligman, M. E. P., & Vaillant, G. E. (1988). Pessimistic explanatory style is a risk factor for physical illness: A thirty-five-year longitudinal study. Journal of Personality and Social Psychology. 55. 23-27. 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Scheier, M. F., & Carver, C. S. (1992). Effects of optimism on psychological and physical well-being: Theoretical overview and empirical update. Cognitive Therapy and Research. 16. 201-228. Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing optimism from neuroticism (and trait anxiety, self-mastery, and self-esteem): A reevaluation of the Life Orientation Test. Journal of Personality and Social Psychology. 67. 1063- 1078. Scheier, M. F., Matthews, K. A., Owens, J. F., Schulz, R., Bridges, M. W., Magovem, G. J., & Carver, C. S. (1999). Optimism and rehospitalization after coronary artery bypass graft surgery. Archives of Internal Medicine. 159. 829-835. Scheier, M. F., Matthews, K. A., Owens, J. F., Schulz, R., Bridges, M. W., Magovem, G. J., Lefebvre, R. C., Abbott, R. A., & Carver, C. S. (1989). Dispositional optimism and recovery from coronary artery bypass surgery: The beneficial effects on physical and psychological well-being. Journal of Personality and Social Psychology. 57. 1024-1040. Scheier, M. F., Weintraub, J. K., & Carver, C. S. (1986). Coping with stress: Divergent strategies of optimists and pessimists. Journal o f Personality and Social Psychology. 51. 1257-1264. Schulz, R., Bookwala, J., Knapp, J. E., Scheier, M., & Williamson, G. M. (1996). Pessimism, age, and cancer mortality. Psychology and Aging. 11. 30-309. Schulz, R., Williamson, G. M., Knapp, J. E., Boolcwala, J., Lave, J. & Fello, M. (1995). The psychological, social, and economic impact of illness among recurrent cancer. Journal of Psychosocial Oncology. 13. 21-45. Schwartzberg, S. S. (1994). Vitality and growth in HIV-infected gay men. Social Science and Medicine. 39. 593-602. Segerstrom, S. C., Taylor, S. E., Kemeny, M. E., & Fahey, J. L. (1998). Optimism is associated with mood, coping, and immune change in response to stress. Journal of Personality and Social Psychology. 74. 1646-1655. Shepperd, J. A., Maroto, J. J., & Pbert, L. A. (1996). Dispositonal optimism as a predictor of health changes among cardiac patients. Journal of Research in Personality. 30. 517-534. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sieber, W. J., Rodin, J., Larson, L., Ortega, S., Cummings, N., Levy, S., Whiteside, T. & Herberman, R. (1992). Modulation of human natural killer cell activity by exposure to uncontrollable stress. Brain. Behavior, and Immunity. 6. 141-156. Siegel, K., & Schrimshaw, E. W. (2000). Perceiving benefits in adversity: stress- related growth in women living with HIV/AIDS. Social Science and Medicine. 51. 1543-1554. Smets, E. M., Visser, M. R., Garssen, B., Frijda, N. H., Oosterveld, P., & Haes, J. C. (1998). Understanding the level of fatigue in cancer patients undergoing radiotherapy. Journal of Psychosomatic Research. 45. 277-293. Smith, T. W., Pope, M. K., Rhodewalt, F., & Poulton, J. L. (1989). Optimism, neuroticism, coping, and symptom reports: An alternative interpretation o f the Life Orientation Test. Journal of Personality and Social Psychology. 56. 640-648. Smith, C. A., Wallston, K. A., Dwyer, K. A. (1995). On babies and bathwater: Disease impact and negative affectivity in the self-reports o f persons with rheumatoid arthritis. Health Psychology. 14. 64-73. Strutton, D., & Lumpkin, J. (1992). Relationship between optimism and coping strategies in the work environment. Psychological Reports. 71. 1179-1186. Sumi, K. (1997). Optimism, social support, stress, and physical and psychological well-being in Japanese women. Psychological Reports. 81. 299-306. Taylor, S.E., & Brown, J.D. (1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin. 103. 193-210. Taylor, S. E., & Gollwitzer, P. M. (1995). Effects of mindset on positive illusions. Journal of Personality and Social Psychology. 69. 213-226. Taylor, S. E., Kemeny, M. E., Aspinwall, L. G., Schneider, S. G., Rodriguez, R., & Herbert, M. (1992). Optimism, coping, psychological distress, and high-risk sexual behavior among men at risk for acquired immunodeficiency syndrome (AIDS). Journal of Personality and Social Psychology. 63. 460-473. Taylor, S.E., Kemeny, M.E., Reed, G. M., Bower, J. E., & Gruenewald, T. L. (2000). Psychological resources, positive illusions, and health. American Psychologist. 55. 99-109. Taylor, S. E., Lichtman, R. R., & Wood, J. V. (1984). Attributions, beliefs about control, and adjustment to breast cancer. Journal of Personality and Social Psychology. 46. 489-502. 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tedeschi, R. G., & Calhoun, L. G. (1995). Trauma & Transformation: Growing in the Aftermath of Suffering. Thousand Oaks, Ca: Sage Publications. Tedeschi, R. G., & Calhoun, L. G. (1996). The posttraumatic growth inventory: Measuring the positive legacy of trauma. Journal of Traumatic Stress. 9. 455-471. Tedeschi, R.G., Park, C.L., & Calhoun, L.G. (Eds.) (1998). Posttraumatic Growth: Positive Changes in the Aftermath of Crises. Mahwah, N.J.: Lawrence Erlbaum Associates. Tennen, H., & Affleck, G. (1998). Personality and transformation in the face of adversity. In Tedeschi, R.G., Park, C.L., & Calhoun, L.G. (Eds.). Posttraumatic Growth: Positive Changes in the Aftermath of Crises. Mahwah, N.J.: Lawrence Erlbaum Associates, 65-98. Tennen, H., Affleck, G., Urrows, S., Higgins, P., & Mendola, R. (1992). Perceiving control, construing benefits, and daily processes in rheumatoid arthritis. Canadian Journal of Behavioral Science. 24. 186-203. Thompson, S. C. (1985). Finding positive meaning in a stressful event and coping. Basic and Applied Social Psychology. 6. 279-295. Tomakowsky, J., Lumley, M. A., Markowitz, N., & Frank, C. (2001). Optimistic explanatory style and dispositional optimism in HIV-infected men. Journal of Psychosomatic Research. 51. 577-587. Williams, R. D., Riels, A. G., & Roper, K. A. (1990). Optimism and distractibility in cardiovascular reactivity. The Psychological Record. 40. 451-457. Yalom, I. D., & Lieberman, M. A. (1991). Bereavement and heightened existential awareness. Psychiatry. 54. 334-345. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Emotional intelligence and smoking risk factors in early adolescents

PDF

Aggression and victimization as predictors of drug use in early adolescence

PDF

Alternatives to condom use and not sharing needles among young injection drug users, their sex partners, and the social network members

PDF

Cigarette smoking, metabolic syndrome, and subclinical atherosclerosis

PDF

Effects of a brief safer sex intervention on provider counseling behaviors, attitudes, and the provider -patient relationship

PDF

Adolescent stress and styles of coping: Predictors and moderators of psychological distress

PDF

A randomized controlled trial of provider independent interventions to increase mammography screening among HMO members who are overdue for a mammogram

PDF

An evaluation of a brief HIV /AIDS prevention intervention using normative feedback to promote risk reduction among sexually active college students

PDF

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

PDF

Correlates of compliance with mammography screening guidelines among low-income Latinas: An exploratory study

PDF

Controlling for biases from measurement errors in health outcomes research: A structural equation modeling approach

PDF

Artery wall injury and LDL-cholesterol in early atherosclerosis: The Los Angeles atherosclerosis study

PDF

Biopsychosocial factors in major depressive disorder

PDF

Cognitive functioning and dementia following cancer: A Swedish twin study

PDF

Anger expression in men with elevated blood pressure: Relationships to assertiveness

PDF

Correlates of smoking behaviour among Egyptian adolescents

PDF

Interaction of dietary fiber and serum cholesterol on early atherosclerosis

PDF

Characterization of neural mechanisms involved in hypoglycemic detection at the portal vein

PDF

Associations and mechanisms among attention deficit hyperactivity disorder symptoms, cognitive functioning, and drinking habits

PDF

An investigation of caregiver sense of control and coping strategies in moderating the effects of caregiving demands

Asset Metadata

Creator Milam, Joel Ellis (author)

Core Title Adaptation to a life -threatening diagnosis: Dispositional optimism and pessimism and posttraumatic growth among patients with HIV

School Graduate School

Degree Doctor of Philosophy

Degree Program Preventive Medicine

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, medicine and surgery,OAI-PMH Harvest,psychology, physiological,psychology, social

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Richardson, Jean (committee chair), [illegible] (committee member), Chou, Chih-Ping (committee member), Meyerowitz, Beth E. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-224150

Unique identifier UC11334792

Identifier 3073815.pdf (filename),usctheses-c16-224150 (legacy record id)

Legacy Identifier 3073815.pdf

Dmrecord 224150

Document Type Dissertation

Rights Milam, Joel Ellis

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA