Conservation of Resources Theory, Perceived Stress, and Chronic Fatigue Syndrome: Outcomes of a Consumer-Driven Rehabilitation Program

Transcription

1 Rehabilitation Psychology 2006, Vol. 51, No. 2, Copyright 2006 by the American Psychological Association /06/$12.00 DOI: / Conservation of Resources Theory, Perceived Stress, and Chronic Fatigue Syndrome: Outcomes of a Consumer-Driven Rehabilitation Program Renee R. Taylor University of Illinois at Chicago Leonard A. Jason DePaul University Yukiko Shiraishi, Michael E. Schoeny, and Jessica Keller University of Illinois at Chicago Objective: To evaluate the impact of a consumer-driven rehabilitation program on perceptions of loss and gain of interpersonal relationships, energy, material objects, work benefits and opportunities, well-being, and experiences of mastery in persons with chronic fatigue syndrome. Study Design: Participants were randomly assigned to a program group (n 23) or a control group (n 24). Outcomes were assessed (a) at baseline, (b) after program participants completed an illness management group, and (c) after they completed one-on-one peer counseling. Setting: A community-based advocacy organization for individuals with disabilities. Interventions: Four months of illness management groups followed by 7 months of one-on-one peer counseling emphasizing goal setting and goal attainment. Main Outcome Measure: The Conservation of Resources Evaluation scale. Results: Significant gains were observed for program participants across all categories of resource gain interpersonal, energy, material, work, well-being, and mastery resources. Effect sizes were moderate to large. Conclusions: Programs in which participatory action research methods are used may have a positive impact on resource acquisition for individuals with chronic fatigue syndrome. Keywords: participatory action research, conservation of resources theory, chronic fatigue syndrome, randomized clinical trial, outcomes Chronic illness can cause significant losses in various key resources, including income, housing, transportation, self-esteem, and social networks, among others. Studies of individuals with chronic illness suggest that these kinds of key losses can have a significant impact on perceived stress and quality of life (Lane & Hobfoll, 1992; Taylor, Kulkarni, & Shiraishi, 2005). For example, losses in well-being, material objects, and interpersonal relationships were associated with poorer perceived quality of life in individuals with chronic fatigue syndrome (CFS; Taylor et al., 2005). To date, few programs for individuals with chronic illness have been developed with the specific intention of offsetting their experiences of loss and increasing their experiences of gain. In the present study, we aimed to examine the effects of one such program on perceptions of gain and loss for individuals with Renee R. Taylor, Yukiko Shiraishi, Michael E. Schoeny, and Jessica Keller, Department of Occupational Therapy, University of Illinois at Chicago; Leonard A. Jason, Department of Psychology, DePaul University. Funding for the Chronic Fatigue Syndrome Empowerment project was provided by the U.S. Department of Education National Institute on Disability and Rehabilitation Research Grant H133G We thank Stefanie Maxwell, Sabrina Johnson, Nancy Carlin, and Heather Eisele for their assistance with study implementation, data collection, and data entry. Correspondence concerning this article should be addressed to Renee R. Taylor, Department of Occupational Therapy, University of Illinois at Chicago, 1919 West Taylor Street, Chicago, IL rtaylor@uic.edu chronic fatigue syndrome using the framework of the conservation of resources theory (CRT; Hobfoll, 1989). CFS is a debilitating and complex disorder characterized by severe fatigue and a number of additional physical and cognitive symptoms (Fukuda et al., 1994). These symptoms include sore throat, swollen lymph nodes, significant difficulties in short-term memory and concentration, unrefreshing sleep, muscle pain, multijoint pain, new-type headaches, and postexertional malaise for more than 24 hr after activity (Fukuda et al., 1994). Psychologists remain just as baffled as physicians and other medical and rehabilitation specialists regarding the true cause and nature of this condition (Mounstephen & Sharpe, 1997). Individuals with CFS tend to report high levels of impairment, particularly in physical functioning, role functioning, social functioning, vitality, and bodily pain (Anderson & Ferrans, 1997; Buchwald, Pearlman, Umali, Schmaling, & Katon, 1996; Hardt et al., 2001; Komaroff et al., 1996; Schweitzer, Kelly, Foran, Terry, & Whiting, 1995). This impairment is not only severe but also, in many cases, enduring. A review of prospective outcome studies in individuals with CFS revealed that substantial recovery occurs in fewer than 10% of cases, and one fourth to one third of individuals with CFS report worsening illness over time (Joyce, Hotopf, & Wessely, 1997). These physical, psychological, and social limitations coexist with experiences of social stigma, public misunderstanding about CFS, and frequently strained relationships with health care providers, friends, and family members (Taylor, Friedberg, & Jason, 2001). We believe that CRT (Hobfoll, 1989, 1998) provides a theoretically grounded means of mapping the type and nature of losses experienced by people with CFS and that it can also serve to 157

2 158 TAYLOR, JASON, SHIRAISHI, SCHOENY, AND KELLER capture the effects of intervention on arresting loss cycles and promoting resource gain. CRT CRT (Hobfoll, 1989, 1998) is an integrated model of stress that can be used to guide interventions and conceptualize program outcomes for individuals with chronic illness. This theory defines the relation between stress and resources in terms of gain and loss. According to the theory, individuals seek to acquire and maintain resources, that is, those objects, personal characteristics, conditions, or energies that are valued by individuals or that serve as a means for attainment of other objects, personal characteristics, conditions, or energies. Object resources include items with a physical presence such as housing, clothing, food, and items indicative of status, such as designer clothing. Personal resources include learned skills and traits such as self-esteem, well-being, and mastery. Condition resources are states that allow access to other types of resources and include health, employment, and marriage. Energy resources, such as money or knowledge, can be exchanged or used in order to acquire other resources. The model proposes that individuals use their resources to acquire additional resources or protect existing ones (Hobfoll, 1989, 1998). For example, an individual entering the health care system expends the energy resources of money and time in order to maintain a condition resource, his or her health. According to CRT (Hobfoll, 1989, 1998), psychological stress occurs in any of three circumstances involving loss or failure to gain resources: (a) when individuals resources are threatened with loss, (b) when individuals resources are lost, and (c) when individuals fail to gain resources following investment of other resources. CRT places loss as the key to perceived stress, as loss is disproportionately weighted in the human experience compared with resource gain (Hobfoll, 1998). Individuals with greater resources are less vulnerable to stress and loss and more capable of exercising their strengths and orchestrating gains. In turn, those who lack resources are more vulnerable to resource loss, and initial loss begets future loss. Acute traumas or negative life events often result in experiences of stress because they produce rapid loss, as in the case of a sudden illness or disasters. Chronically stressful conditions gnaw away at resources, even chiseling away strong resource reservoirs. The circumstance of having a chronic illness can be associated with a number of acute and chronic resource losses. Studies of CRT, as applied to individuals with chronic illness, have illustrated that, in the absence of intervention, loss is more significant than gain (Lane & Hobfoll, 1992; Taylor et al., 2005). One study of individuals with a severe chronic breathing disorder revealed that illness symptoms and consequent losses were associated with angry feelings that were linked to angry behavior that resulted in increased anger in close others (Lane & Hobfoll, 1992). The authors concluded that chronic stress can perpetuate additional losses in characteristics such as well-being, self-esteem, and interpersonal relationships, predisposing people to an ongoing cycle of chronic stress and perceived loss (Lane & Hobfoll, 1992). A second study of individuals with CFS (Taylor et al., 2005) found that overall perception of loss and loss of well-being, material objects, and interpersonal relationships were significantly associated with lower quality of life. Rehabilitation and CFS Presently, there is no widely accepted treatment for CFS, and only a small number of comprehensive, interdisciplinary rehabilitation programs exist (Essame, Phelan, Aggett, & White, 1998; Lim & Lubitz, 2002; Marlin, Anchel, Gibson, Goldberg, & Swinton, 1998; Pemberton, Hatcher, Stanley, & House, 1994). A growing number of research studies have been initiated to test the efficacy of such integrative rehabilitation programs, and there is preliminary evidence that these programs improve specific aspects or consequences of the syndrome, such as fatigue severity (Chalder, Wallace, & Wessely, 1997), physical and occupational functioning (Essame et al., 1998; Marlin et al., 1998; Pemberton et al., 1994; Sadlier, Phil, Evans, Phillips, & Broad, 2000), and psychological distress (Chalder et al., 1997; Soderberg & Evengard, 2001). In the present study, we aimed to examine the effects of a consumer-driven rehabilitation program on resource gain and loss for individuals with CFS using the framework of CRT (Hobfoll, 1989). The Chronic Fatigue Syndrome Empowerment project (Taylor & Jason, 2002) is an integrative, participatory rehabilitation program designed to improve resource acquisition and quality of life for individuals with CFS. Following an initial focus group, participants in the program received 4 months of illness management groups covering a number of topics relevant to the management of CFS (Part 1) followed by a 7-month one-on-one selfadvocacy training period during which they continued to set and attain goals and also learn and practice strategies to obtain needed resources and services (Part 2). Aims and Hypotheses The central aim of this study was to evaluate the cumulative effects of this two-part program on participants ability to strengthen existing resources and arrest loss spirals that may be associated with having CFS. Individuals with CFS were randomly assigned to either an immediate-program group or a delayedprogram control group and evaluated concurrently at three time points (baseline, postgroup, and post one-on-one). The primary hypothesis was that the program would lead to increases in overall resource gain and decreases in overall resource loss for program participants over time (i.e., Program Condition Time interaction). A secondary hypothesis was that the program would lead to increases in gain and decreases in loss of the following resource subdomains: self-esteem, well-being, mastery, work, material objects, energy, interpersonal relationships, and family relationships. Method Design Participants were randomly assigned to either an immediate-program condition or a delayed-program control condition. Those in the immediateprogram condition first participated in eight sessions of an illness management group, occurring every other week over a period of 4 months. Following a postgroup assessment during a 1-month break period, program participants then completed 7 months of one-on-one peer counseling, followed by another assessment. Delayed-program controls only completed

3 CHRONIC FATIGUE SYNDROME 159 the assessments during this 12-month period. Please refer to Figure 1 for a depiction of the flow of participants through various stages of the study design. The Conservation of Resources Evaluation (COR-E; Hobfoll & Lilly, 1993) outcomes measure was administered by research assistants not involved in the intervention and not informed of group assignment. These assistants were trained on COR-E administration by Renee R. Taylor. Measures were administered uniformly to both experimental and control participants at the same three time points throughout the intervention: (a) at baseline, (b) within 1 month after participants in the immediate-program condition completed the group phase, and (c) within 1 month after participants in the immediate-program condition completed the one-on-one phase. Participants Fifty-two women and men were recruited between October 1, 2000 and December 1, 2000 from the following sources: (a) local CFS self-help organizations and Chicago-area physicians specializing in the treatment of people with CFS; and (b) advertisements posted in CFS newsletters, Chicago-area newspapers, on CFS Web sites and Listservs, and on a local cable TV station. Prospective participants provided informed consent by receiving and signing a consent form, receiving a follow-up phone call to clarify all procedures in the study and answer any questions, and returning the form in the mail. All recruited adults underwent a screening process to confirm their self-reported diagnosis of CFS. This process involved four steps: (a) participants completed the Chronic Fatigue Syndrome Screening questionnaire (Jason et al., 1997) to evaluate the presence, frequency, and severity of CFS symptoms according to the present U.S. diagnostic criteria (Fukuda et al., 1994); (b) participants then completed a semistructured psychiatric interview, The Structured Clinical Interview for the DSM IV (SCID) (First, Spitzer, Gibbon, & Williams, 1995), administered by a licensed clinical psychologist to rule out psychiatric conditions that would exclude an individual from a CFS diagnosis according to the U.S. criteria (Fukuda et al., 1994); (c) collection of past medical records documenting a diagnosis of CFS by a physician; and (d) independent physician review of results from the Chronic Fatigue Syndrome Screening questionnaire, the psychiatric interview, and the medical records to determine whether the potential participants met CFS criteria (Fukuda et al., 1994). On the basis of the screening process, a diagnosis of CFS, according to the present international research diagnostic criteria (Fukuda et al., 1994), was confirmed by an independent physician reviewer for 50 individuals. Two participants were excluded: one had an exclusionary medical condition (untreated hyperthyroidism) and the other did not meet the Fukuda et al. (1994) requirement of four or more associated symptoms of CFS. On the basis of findings from the SCID, none of the participants met diagnostic criteria for any of the disorders that would exclude a person from a diagnosis of CFS. Exclusionary diagnoses (Fukuda et al., 1994) included psychotic disorders, bipolar disorders, dementias, organic brain disorders, major depression with melancholic or psychotic features, any substance abuse within 2 years before CFS onset, anorexia nervosa, and bulimia. Of the remaining 50 individuals, 3 refused random assignment, and they were not included in the statistical analyses of outcomes. Of these, 1 participated in the immediate-program group and the other 2 participated in the delayed-program group. Thus, the final sample consisted of 47 individuals with CFS. Twenty-three individuals were randomly assigned to the immediate-program group, and 24 were assigned to the delayed-program control group. Random assignment was completed by a research assistant proficient in statistics using the SPSS for Windows Version 11.0 random variable uniform function. Using this function, participants were assigned one of two numeric codes and were assigned to groups by Renee R. Taylor on the basis of that code. Because controls waited 1 year to receive the program (after program participants completed the program), participants were aware of their group assignment. Informing participants of their group assignment was required by the internal review board of the supporting university for the protection of human participants. Program adherence was good, and there were no dropouts during the program period. Table 1 presents sociodemographic characteristics of the sample. Of the 47 participants, most were women of middle age, of middle socioeconomic status, and single (either never married, separated, widowed, or divorced). Most identified as Caucasian, and most were currently unemployed. The revised scoring rules for Hollingshead s (1975) scale, developed and validated by Wasser (1991), were used to estimate socioeconomic status from data on occupation and education for each participant, resulting in the following three categories: (a) low, (b) middle, and (c) high. Chi-square and t-test analyses revealed no significant differences in sociodemographic characteristics between program participants and controls. On the basis of findings from the SCID, 41% of the 47 participants included in the study received at least one current nonexclusionary psychiatric diagnosis. Nineteen percent of participants were diagnosed with a current depressive disorder (nonpsychotic and nonmelancholic), and 22% Figure 1. Flow of participants through stages of the study design.

4 160 TAYLOR, JASON, SHIRAISHI, SCHOENY, AND KELLER Table 1 Descriptive Data on Sociodemographic Characteristics of Participants According to Program Condition Characteristic were diagnosed with a current anxiety disorder (including posttraumatic stress disorder). Program participants and controls did not differ in rates of current psychiatric disorder, , p.05. Program Description Program participants (n 23) Comparison group Controls (n 24) f % f % Work status Full time Part time Not working Socioeconomic status Low Middle High Gender Female Male Marital status Never married Married Divorced, widowed, separated Ethnicity Minority Nonminority Age M SD Note. f frequency. The program took place within a center for independent living, a community-based advocacy organization for individuals with disabilities. This center employed two peer counselors with CFS on a part-time basis, and these individuals served as the primary program providers. Both peer counselors had been employed in full-time human service positions prior to CFS onset, one as a nurse and the other as a police officer. Both peer counselors had experience with group leadership and had served in other lay counseling roles. These peer counselors were supervised closely by a licensed clinical psychologist throughout the program. They co-led the illness management groups with the psychologist and conducted the oneon-one peer counseling with participants on their own, with supervision from the psychologist taking place on a weekly basis. This project is unique in the study of CFS in that it applies the principles of participatory action research to its design, implementation, evaluation, and dissemination. Participatory action research involves participants in an egalitarian partnership with researchers. In essence, it seeks to enable people to reestablish power and control in their own lives and to realize their power as a member of a collective community. Formal boundaries between traditional roles (e.g., researcher, consumer, service provider) are reduced or eliminated, and anyone who participates may assume a variety of roles and responsibilities in the research and action process. Accordingly, participatory action research provides a framework in which people with disabilities can take an active role in designing and conducting research (Balcazar, Keys, Kaplan, & Suarez-Balcazar, 1998). For the present study, the structure and logistical elements of the program and the research process were developed conjointly by members of the local self-help organization serving individuals with CFS, staff of the center for independent living, and Renee R. Taylor and Leonard A. Jason. Knowledge of the benefits and wide-ranging impact of participatory action research is growing within the social sciences and rehabilitation communities (e.g., mandating of the use of this approach by some federal granting agencies). However, randomized clinical trials in which this approach is used remain a contemporary rarity in the social sciences literature. Program participants first participated in eight sessions of an illness management group, occurring every other week over a period of 4 months. This group was co-led by a peer counselor and Renee R. Taylor. The first hour of each group session consisted of individual check-ins and reporting on self-monitored goal attainment. The second hour of each group session consisted of an educational lecture and discussion of self-selected, CFSrelevant topics. Similar group structures have been used with individuals with other types of chronic conditions, such as rheumatoid arthritis (Lorig et al., 1999). In our program, group topics included activity pacing using the envelope theory (Taylor et al., 2001), cognitive coping skills training, relaxation and meditation training, employment issues and economic selfsufficiency, personal relationships, traditional and complementary medical approaches, and nutritional approaches. More information about the content of these groups is provided by Taylor and Jason (2002). More information about the types of goals that participants set for themselves in the program is provided by Query and Taylor (2005). Following the illness management group sessions, program participants received 7 months of peer counseling, consisting of self-advocacy training, continued monitoring of goal attainment, and ongoing case coordination services that were individually tailored to the needs of each participant. Each counseling session was provided by one of the two peer counselors assigned to work with the given participant. Each participant received a weekly 30-min telephone call at home at a mutually negotiated time, during which the participant updated the peer counselor on their progress on program goals. The same goal attainment form used during Part 1 of the program was mailed to the participant for use during Part 2, and the same educational topics covered in Part 1 were emphasized in Part 2. However, the degree to which a given topic from Part 1 was emphasized during Part 2 varied according to the individual needs of the participant. The participant and the peer counselor problem solved about ways to overcome obstacles to goal attainment. When indicated, referrals were made to a wide range of public assistance programs and other community-based service providers to facilitate goal attainment. Participants received an average of 23 sessions during the 7-month period (SD 5.34). Measures Screening measures. The Chronic Fatigue Syndrome Screening questionnaire (Jason et al., 1997) is a combination of existing and new measures, including (a) sociodemographic characteristics; (b) the Fatigue Scale (Chalder et al., 1993); (c) the Chronic Fatigue Syndrome Symptom Rating Form; and (d) questions assessing symptoms of CFS, quality, and duration of fatigue. The other questions assessing symptoms of CFS, quality, and duration of fatigue are scored according to the Fukuda et al. (1994) criteria. Reliability for the demographic questions has been determined to be high (interrater agreement 87% 100%, ; test retest agreement 92% 100%, ) (Jason et al., 1997). Similarly, the total score on the Fatigue Scale has demonstrated high interrater reliability (r.81) and excellent test retest reliability (r.95) (Jason et al., 1997). Reliability for the symptom questions has also been estimated to be high (interrater agreement 76% 92%, ; test retest agreement 76% 92%, ) (Jason et al., 1997). The SCID (First et al., 1995) is a professionally administered, semistructured psychiatric interview that was used to diagnose Axis I psychiatric disorders according to DSM IV criteria. The SCID is appropriate for use with individuals with CFS because it allows for clinical judgment in the assignment of symptoms to psychiatric or medical categories, a crucial distinction in diagnosing CFS (Friedberg & Jason, 1998; Taylor & Jason,

5 CHRONIC FATIGUE SYNDROME ). Symptoms within each diagnostic category are scored as either absent, subthreshold, or present, and all symptoms that are present are counted toward the diagnostic tally as it conforms to DSM IV criteria. A psychodiagnostic study (Taylor & Jason, 1998) validated the use of the SCID in a sample of individuals with CFS. Although in a large-scale multisite study, Williams and associates (1992) found the SCID to have adequate test retest reliability, these same researchers argued that the reliability of the SCID is highly dependent on the clinical acuity and training of the interviewer. In the present study, Renee R. Taylor, who had been administering the SCID in research studies for approximately 7 years, conducted all of the SCID interviews. Accuracy of diagnoses made using the SCID was checked against the participant s medical records and self-reported psychiatric history. Outcome measures. The COR-E (Hobfoll & Lilly, 1993) is a 74-item inventory that measures the extent of the past month s gain and loss of resources described by CRT: objects, conditions, personal characteristics, and energies (Hobfoll, 1989, 1998). The scale is a valid and reliable measure of stress, as it is defined by resource gain and loss (Ennis, Hobfoll, & Schroder, 2000; Freedy, Saladin, Kilpatrick, Resnick, & Saunders, 1994; Ironson et al., 1997). The COR-E has a Loss scale and a Gain scale, each containing the same 74 items. Items in the Loss scale are preceded by the question: To what extent have you lost the following resource during the past month. Items in the Gain scale are preceded by the question: To what extent have you gained the following resources during the past month. Participants rated the amount of loss, ranging from 0 (no loss) to7(a great degree of loss), and the amount of gain in the past month, ranging from 0 (no gain) to7(a great degree of gain), for each resource item. Responses were summed for the two scales to yield scores reflecting total resource loss and total resource gain (each score ranged from 0 to 518). In addition, gain and loss scores for the following subscales were summed in the same manner: Self-Esteem, Well-Being, Mastery, Work, Material Objects, Energy, Interpersonal Relationships, and Family Relationships. Statistical Analyses Given the nested design of this study (i.e., observations nested within individuals), linear growth models were estimated, comparing program and control conditions in terms of resource gain and loss using random effects regression analyses. This approach was selected over a repeated measures analysis of variance (ANOVA) because it more accurately models the effect of interest, which is change over time according to condition. More specifically, random effects regression analysis allowed us to measure the cumulative effects of both phases of the intervention on program participants over time. We were also interested in comparing these effects against the rate of natural change in control participants who did not receive the program. Although a repeated measures ANOVA would have been an appropriate alternative for analyzing these data, another strength of random effects regression analysis is its greater flexibility regarding the treatment of missing data. All analyses were performed using hierarchical linear modeling (HLM 5; Raudenbush, Bryk, Cheong, & Congdon, 2000). These models consisted of Level 1 intercepts (set to the final wave of data) and slopes (i.e., changes over time), each of which is allowed to vary across individuals (i.e., has a random component). Program condition was treated as a fixed, Level 2 effect. These random effects regression models were used to test the cumulative effects of both the group phase and the one-on-one phase of the program over time in program participants as compared with controls. Testing this nested model allowed examination of the hypothesis that participants in the program would demonstrate significantly greater change in outcomes over time (i.e., Program Condition Time interaction), resulting in significantly higher resource gain and significantly lower resource loss at the end of the intervention phase (i.e., program condition main effect). Separate models for overall resource gain and loss and separate models for the eight subscales of resource gain and loss (selfesteem, well-being, mastery, work, material objects, energy, interpersonal relationships, and family relationships) were tested. In each model, the effects of time (Level 1) and condition (Level 2) were tested as predictors, and the various categories of resource gain and loss served as the outcome variables. The intercept was centered at Wave 3 in order to most conservatively test the outcomes of the program over time. To test for the possibility that participants baseline fatigue severity may have affected program outcomes, this factor was initially included as a covariate at Level 2. Findings remained unchanged, so the covariate was removed from the model. To aid in the determination of the clinical significance of observed statistical differences, effect sizes (Cohen s d) were calculated with 95% confidence intervals (Rosenthal & Rosnow, 1991). Effect size measures the magnitude of treatment effects independent of sample size. Results Overall Resource Gain and Loss There was a significant main effect of the program for overall resource gain (b 39.50, SE 17.02), t(45) 2.32, p.05, with a moderate effect size (Cohen s d 0.69). A significant Time Condition interaction effect on overall resource gain was also found (b 35.00, SE 10.71), t(45) 3.27, p.05, for which the effect size was large (Cohen s d 0.97). As indicated in Table 2, program participants reported significantly more resource gain over time relative to those in the control group, who reported a decline in resource gain. Table 2 also indicates that perceived overall resource gain was highest following the group phase at Assessment 2. There were no significant main or interaction effects of the program on overall resource loss. Resource Subdomains Table 2 also presents means and standard deviations for all resource subdomains. Analysis of self-esteem showed a significant main effect of condition (b 5.63, SE 2.62), t(45) 2.15, p.05, d 0.78, as well as a significant Condition Time interaction effect (b 4.38, SE 1.51), t(45) 2.89, p.01, d Individuals in the program group reported a significant increase in self-esteem resource gain over time, whereas those in the control group reported decline. With reference to well-being, findings revealed a significant Time Condition interaction (b 3.85, SE 1.39), t(45) 2.77, p.01, d 0.83, such that individuals in the program group reported a significant increase in well-being over time as compared with controls. Analysis of gain in mastery showed a significant main effect of condition (b 6.38, SE 2.97), t(45) 2.15, p.05, d 0.64, and a significant Condition Time interaction (b 4.37, SE 1.73), t(45) 2.52, p.05, d Compared with controls, individuals in the program group reported a significant increase in mastery resource gain over time. Similarly, analysis of work gain showed a significant Condition Time interaction effect (b 3.91, SE 1.71), t(45) 2.29, p.05, d Individuals in the program group reported a significant increase in work resource gain over time, and controls reported a decline. With respect to material objects, findings also revealed a significant Time Condition interaction (b 4.43, SE 1.89), t(45) 2.34, p.05, d 0.79, such that individuals in the program group reported significant increase in material resource gain over time, whereas those in the control group reported decline. Analysis of energy revealed a significant main

6 162 TAYLOR, JASON, SHIRAISHI, SCHOENY, AND KELLER Table 2 Means and Standard Deviations for COR-E Scales for Participants With Chronic Fatigue Syndrome According to Program Condition Variable Program participants (n 23) Comparison group Controls (n 24) Program participants (n 23) Comparison group Controls (n 24) M SD M SD Variable M SD M SD Primary domain Overall resource gain a * Assessment Assessment Assessment Overall resource loss Assessment Assessment Assessment Resource Subdomain Self-esteem gain a ** Assessment Assessment Assessment Self-esteem loss Assessment Assessment Assessment Well-being gain a ** Assessment Assessment Assessment Well-being loss Assessment Assessment Assessment Mastery gain a * Assessment Assessment Assessment Mastery loss Assessment Assessment Assessment Work gain a * Assessment Assessment Assessment Note. COR-E Conservation of Resources Evaluation. a Indicates a significant Time Condition interaction. * p.05. ** p.01. Resource Subdomain (continued) Work loss Assessment Assessment Assessment Material gain a * Assessment Assessment Assessment Material loss Assessment Assessment Assessment Energy gain a * Assessment Assessment Assessment Energy loss Assessment Assessment Assessment Interpersonal gain a * Assessment Assessment Assessment Interpersonal loss Assessment Assessment Assessment Family gain Assessment Assessment Assessment Family loss Assessment Assessment Assessment effect of condition on energy resource gain (b 6.77, SE 2.49), t(45) 2.72, p.05, d 0.81, and a significant Condition Time interaction effect (b 5.45, SE 1.96), t(45) 2.78, p.01, d Compared with controls decline in energy resource gain, individuals in the program group reported a significant increase in energy resource gain over time. Likewise, analysis of interpersonal relationships revealed a significant Time Condition interaction (b 4.45, SE 2.02), t(45) 2.21, p.05, d Program participants reported a significant increase in interpersonal resource gain, whereas those in the control group demonstrated a decline. All perceived gains in self-esteem, well-being, mastery, work, material objects, energy, and interpersonal relationships were greatest following the group phase at Assessment 2 (see Table 2). Analysis of gain and loss of family relationships revealed no significant main or interaction effects. In addition, there were no significant effects of resource loss in the remaining subdomains of self-esteem, well-being, mastery, work, material objects, energy, and interpersonal relationships. Discussion This study constitutes the first randomized clinical trial measuring the effects of a participatory community-based rehabilitation

7 CHRONIC FATIGUE SYNDROME 163 program on perceived resource acquisition for individuals with chronic illness. Findings from this study indicate that communitybased rehabilitation programs, such as the one described herein, may lead to increased resource gain in major areas of life for individuals with CFS. Specifically, participation in the program was linked to a perception of greater resource gain overall and in a number of discrete areas self-esteem, well-being, mastery, work, material goods, energy, and interpersonal relationships. According to CRT (Hobfoll, 1989, 1998), these are valued commodities that can serve as a means of reducing and preventing stress and improving quality of life. Researchers have argued that individuals with CFS have low self-esteem. One well-designed study comparing individuals with CFS, other chronic illnesses, and healthy controls found significantly lower underlying self-esteem among individuals with CFS (Creswell & Chalder, 2002). Improvement in self-esteem is not only critical, in itself, for an individual coping with a condition as debilitating as CFS, but it is also important in terms of its potential impact on the development of comorbid mood disorders (Moss- Morris & Petrie, 2001). Adequate self-esteem is also necessary for engagement in productive interpersonal relationships. Similarly, an improved sense of mastery is vital to engagement in work (Kielhofner, 2002), and work is necessary for retention and acquisition of material goods necessary for survival. Most individuals with CFS have had to reduce their work hours, change occupations, or quit work entirely because of CFS (Taylor & Kielhofner, 2005). Interventions like this one that aim to improve perceived mastery and the ability to engage in work are much needed within the CFS community. Findings from this study also suggest that the program improved the perception of energy among participants. According to CRT, energy is a resource that can be exchanged or used in order to acquire other resources. This outcome points to the possibility that the program allowed participants to build on their existing strengths so that they could improve their functioning in other life areas. Family was the only area in which participants did not perceive a significant gain in resources as a result of the program. It may have been more difficult for participants to alter circumstances involving family relationships, particularly when the program was not specifically developed as a family intervention. Though it was not possible to determine the exact components of the program that led to the perception of increased resource gain, findings suggest that program participants demonstrated greatest improvement at Assessment 2, which occurred after the eight sessions of the illness management group. It is unclear why gains were more modest following the one-on-one peer counseling phase of the program. One possibility involves the added effect of positive group identity resulting from contact with and social support from peers facing similar circumstances, both of which occur within a group format. Another possibility may involve the fact that the groups were co-led by a psychologist. Added information offered by a psychologist with wide-ranging knowledge about both CFS management approaches and available resources in the community may have served to increase participants perception of gain. An alternative explanation is that the attenuation in resource gains observed at Assessment 3 may reflect the fact that the end of the program represented a major loss for the participants, and one that contributed to reduction in scores on many scales. It is also possible that only sharing progress on goals with an individual peer counselor had less impact than problem solving and receiving support for goal attainment in a group format. Our study has certain limitations. Given that participants knew the objectives of the program and the group to which they were assigned, experimental expectancy effects may have contributed to some of the improvements in the program group. In particular, improvements in self-esteem, well-being, and energy among program participants may, in part, have been because of their awareness that the program was designed to help in these areas. Conversely, the declines observed over time in the control group may, in part, have been because of the fact that control participants knew that they would not be receiving the program for a year. This violation of pure random assignment was imposed for the protection of human participants and was therefore unavoidable. Another limitation of this study was the small sample size, which prevented us from performing a components analysis. Such an analysis would have allowed us to determine the independent effectiveness of each program phase (group intervention vs. oneon-one intervention). Future studies should tease apart not only the effectiveness of each approach but also how these two approaches may affect each other to influence the effectiveness of integrative rehabilitation programs for individuals with CFS. A further limitation of this study is that it does not provide data on outcomes of the control group, who completed the program 1 year following the program participants. Future studies are necessary to determine whether the outcomes observed in program participants can be replicated in the delayed entry control group. The program did not appear to affect resource loss. Instead, perception of resource loss seemed to decrease over time for both program participants and controls. This finding contradicts CRT because the theory suggests that those who have a reduction in resource gain are at risk for an increased loss of resources. One explanation for this finding involves the fact that the baseline assessment was conducted before participants were formally accepted into the program and randomly assigned to groups. Thus, not knowing whether one was going to be accepted into the program may have served to increase perceptions of resource loss at the beginning of the study. Before the program began and once they were accepted into the program, participants were then randomly assigned to groups. Knowledge that one would receive the program, regardless of whether immediately or 1-year later (for controls), may have attenuated perception of resource loss at assessment time points 2 and 3 for both groups. It is also possible that, despite being asked to report losses occurring only in the past month, participants reported greatest perceived loss at baseline because being asked about loss can prompt a need for witnessing and validation. Moreover, some participants may have incorrectly reflected on a more cumulative time period, such as the time since they were first ill with CFS, rather than on the past month as instructed. In later follow-ups, participants may have more accurately reported only on the past-month time period. Although these limitations pose some risk to the validity of the findings, it remains unlikely that participants perceptions of resource loss and gain would be equally contaminated across all subdomains of resource loss and gain. It is more likely that participants knowledge of group assignment affected their perception of self-esteem, well-being, and energy. However, it is less likely that knowledge of group assignment affected their percep-

8 164 TAYLOR, JASON, SHIRAISHI, SCHOENY, AND KELLER tion of improvement in the areas of work, mastery, interpersonal relationships, and material gain. Given these possibilities, even a conservative interpretation of the findings indicates that participants perceived a number of benefits and gains related to their involvement in the program. Findings from this study carry a number of implications for community-based psychological interventions. First, they illustrate that integrative, community-based rehabilitation programs that are peer-facilitated and emphasize psychoeducation, goal setting, and self-advocacy can have a modest but clinically significant positive effect on perceptions of resource gain for individuals with CFS. Generally speaking, participatory, community-based models of intervention may offer a viable alternative to more traditional approaches to therapy, which typically focus on correcting deficits or improving functioning in a single area. In participatory approaches, mental health professionals can assume a number of interrelated roles. These include assisting clients to identify their own treatment goals and self-monitor their progress; inviting clients with particular strengths or experience to co-lead a rehabilitation program or serve as peer counselors with supervision; and establishing an active working relationship with local centers for independent living so that clients can learn principles of selfadvocacy, disability rights, and gain access to community-based services and resources that facilitate independent living. All of these actions were carried out effectively by the peer counselors in the present study. The use of peer counselors in the rehabilitation process, if appropriately trained and supervised on an ongoing basis, can be a powerful instrument for change. Anecdotally, participants in the program reported that peer counselors offered a unique level of empathy and shared an abundance of relevant real-world experiential information; sharing their own experiences with applying aspects of the illness management groups to their own lives; sharing their personal struggles with CFS; helping clients navigate social systems of care (i.e., applying for social security disability benefits; advocating for reasonable accommodation within the workplace; or applying for assistance to receive food delivery, electricity, gas, or housing). Psychologists in clinical practice and their clients may benefit from increased expertise, empathy, and knowledge acquired through direct collaboration with individuals with CFS who are supervised to function in peer-counseling roles as members of a larger interdisciplinary treatment team. In summary, findings from this study suggest that integrative, peerfacilitated rehabilitation programs that focus on goal setting and education may lead to increased resource acquisition for individuals with CFS. The extent to which findings can be generalized to a clinical setting is unclear, but, given the effect sizes observed, it is likely that these findings are substantially generalizable to community-based centers for independent living similar to the one described herein. References Anderson, J. S., & Ferrans, C. E. (1997). The quality of life of persons with chronic fatigue syndrome. Journal of Nervous and Mental Diseases, 185, Balcazar, F. E., Keys, C. B., Kaplan, D. L., & Suarez-Balcazar, Y. (1998). Participatory action research and people with disabilities: Principles and challenges. Canadian Journal of Rehabilitation, 12, Buchwald, D., Pearlman, T., Umali, J., Schmaling, K., & Katon, W. (1996). Functional status in patients with chronic fatigue syndrome, other fatiguing illnesses, and healthy individuals. American Journal of Medicine, 101, Chalder, T., Berelowitz, G., Pawlikowska, T., Watts, L., Wessely, S., Wright, D., et al. (1993). Development of a fatigue scale. Journal of Psychosomatic Medicine, 37, Chalder, T., Wallace, P., & Wessely, S. (1997). Self-help treatment of chronic fatigue in the community: A randomized controlled trial. British Journal of Health Psychology, 2, Creswell, C., & Chalder, T. (2002). Underlying self-esteem in chronic fatigue syndrome. Journal of Psychosomatic Research, 53, Ennis, N., Hobfoll, S. E., & Schroder, K. E. E. (2000). Money doesn t talk, it swears: How economic stress and resistance resources impact innercity women s depressive mood. American Journal of Community Psychology, 28, Essame, C. S., Phelan, S., Aggett, P., & White, P. D. (1998). Pilot study of a multidisciplinary inpatient rehabilitation of severely incapacitated patients with the chronic fatigue syndrome. Journal of Chronic Fatigue Syndrome, 4, First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1995). Structured Clinical Interview for DSM IV Axis I Disorders Patient edition. New York: Biometrics Research Department. Freedy, J. R., Saladin, M. E., Kilpatrick, D. G., Resnick, H. S., & Saunders, B. E. (1994). Understanding acute psychological distress following natural disaster. Journal of Traumatic Stress, 7, Freedy, J. R., Shaw, D., Jarrel, M. P., & Master, C. R. (1992). Toward an understanding of the psychological impact of disasters. Journal of Traumatic Stress, 5, Friedberg, F., & Jason, L. A. (1998). Understanding chronic fatigue syndrome: An empirical guide to assessment and treatment. Washington, DC: American Psychological Association. Fukuda, K., Strauss, S. E., Hickie, I., Sharpe, M. C., Dobbins, J. G., & Komaroff, A. (1994). The chronic fatigue syndrome: A comprehensive approach to its definition and study. Annals of Internal Medicine, 121, Hardt, J., Buchwald, D., Wilks, D., Sharpe, M., Nix, W. A., & Egle, U. T. (2001). Health-related quality of life in patients with chronic fatigue syndrome: An international study. Journal of Psychosomatic Research, 51, Hobfoll, S. E. (1989). Conservation of resources. A new attempt at conceptualizing stress. American Psychologist, 44, Hobfoll, S. E. (1998). Stress, culture, and community: The psychology and philosophy of stress. New York: Plenum Press. Hobfoll, S. E., & Lilly, R. S. (1993). Resource conservation as a strategy for community psychology. Journal of Community Psychology, 21, Hollingshead, A. B. (1975). Four-factor index of social status. Unpublished manuscript, Yale University. Ironson, G., Wynings, C., Schneiderman, N., Baum, A., Rodriguez, M., Greenwood, D., et al. (1997). Posttraumatic stress symptoms, intrusive thoughts, loss, and immune function after Hurricane Andrew. Psychosomatic Medicine, 59, Jason, L. A., Ropacki, M. T., Santoro, N. B., Richman, J. A., Heatherly, W., Taylor, R., et al. (1997). A screening scale for chronic fatigue syndrome: Reliability and validity. Journal of Chronic Fatigue Syndrome, 3, Joyce, J., Hotopf, M., & Wessely, S. (1997). The prognosis of chronic fatigue and chronic fatigue syndrome: A systematic review. Quarterly Journal of Medicine, 90, Kielhofner, G. (2002). A model of human occupation: Theory and application (3rd ed.). Philadelphia: Lippincott Williams & Wilkins. Komaroff, A. L., Fagioli, L. R., Doolittle, T. H., Gandek, B., Gleit, M. A., Guerrierio, R. T., et al. (1996). Health status in patients with chronic