Testing Self-Efficacy as a Pathway That Supports Self-Care Among Family Caregivers in a Psychoeducational Intervention

Journal of Family Social Work, 13:149 162, 2010 Copyright # Taylor & Francis Group, LLC ISSN: 1052-2158 print=1540-4072 online DOI: 10.1080/10522150903487107 Testing Self-Efficacy as a Pathway That Supports Self-Care Among Family Caregivers in a Psychoeducational Intervention MARIE Y. SAVUNDRANAYAGAM University of Wisconsin Milwaukee, Milwaukee, Wisconsin MARY BRINTNALL-PETERSON University of Wisconsin-Extension, Madison, Wisconsin This study investigated the extent to which a psychoeducational intervention supports family-centered care by influencing health risk and self-care behaviors of caregivers of individuals with Alzheimer s disease (N ¼ 325). Moreover, this study investigated the extent to which changes in self-efficacy explained changes in health risk and self-care behaviors. Data were analyzed using repeated measures and multivariate analysis of variance and multiple regression. Qualitative written accounts of the impact of the intervention augmented the quantitative findings. The study s findings revealed that family caregivers experienced reductions in health risk behaviors and improvements in exercise, stress management, and relaxation activities as a result of participating in the psychoeducational intervention. Improvement in selfefficacy was linked with reductions in health risk behaviors and increased involvement in stress management and relaxation activities. Implications are discussed in terms of the need to understand the mechanisms by which interventions influence family caregivers and the role these mechanisms play in supporting family-centered care. The authors acknowledge financial support from grants to M. Y. Savundranayagam from the Hartford Foundation s Geriatric Social Work Faculty Scholars Program and the Center on Age and Community at the University of Wisconsin Milwaukee. Address correspondence to Marie Y. Savundranayagam, Helen Bader School of Social Welfare, University of Wisconsin Milwaukee, P. O. Box 786, Milwaukee, WI 53201. E-mail: maries@uwm.edu 149

150 M. Y. Savundranayagam and M. Brintnall-Peterson KEYWORDS Alzheimer s disease, family caregivers, family-centered care, psychoeducational interventions, self-care, self-efficacy Family members caring for an individual with Alzheimer s disease (AD) face multiple challenges due to the various cognitive, functional, behavioral, and communication impairments that are symptoms of the illness (Bédard, Pedlar, Martin, Mallott, & Stones, 2000; Pinquart & Sorenson, 2003; Savundranayagam, Hummert, & Montgomery, 2005). The rigorous care demands and exhaustive stressors associated with AD require exceptional vigilance. They can threaten the familial relationships and negatively affect the caregiver s mental and physical health. Caregivers who provide care to persons with AD reported higher levels of anxiety, stress, and depression than those caring for persons with physical disabilities and cancer (Clipp & George, 1993; Hooker, Monahan, Bowman, Frazier, & Shifren, 1998; Ory, Hoffman, Yee, Tennstedt, & Schulz, 1999). Mental health consequences have also been linked with elevated stress hormones (Vitaliano, Zhang, & Scanlan, 2003) and increased risk for death among caregivers (Schulz & Beach, 1999). Negative health outcomes associated with caregiving have been linked with reduced self-care and greater numbers of health risk behaviors, such as poor diet and exercise and not getting enough rest (Gallant & Connell, 1997; Vitaliano et al., 2002). Therefore, a familycentered approach to the care of individuals with AD requires a close examination of the benefits of self-care and a strong effort to reinforce this information with family caregivers. More importantly, it is vital that family caregivers are educated, empowered, and supported in their efforts to engage in positive self-care behaviors as they continue in their caregiving role. OVERVIEW OF POWERFUL TOOLS FOR CAREGIVING A psychoeducational program called Powerful Tools for Caregiving (PTC) was developed to address the issue of self-care for informal caregivers (including family and friends) of individuals with chronic illnesses. PTC runs for six consecutive weeks, with each session lasting 2½ hours. The structure of PTC is based on the Chronic Disease Self-Management Program (Lorig & Holman, 2003; Lorig et al., 1996), a patient education program that relies heavily on developing and building the participants self-efficacy. PTC uses four strategies to enhance self-efficacy: skills mastery, modeling, reinterpretation of feelings and attitudes about caregiving, and persuasion (Boise, Congleton, & Shannon, 2005). Using these strategies, PTC teaches caregivers to manage their emotions, engage in self-care behaviors (e.g., take time to relax), and communicate assertively (Schmall, Cleland, & Sturdevant, 2000).

Self-Efficacy and Self-Care Among Family Caregivers 151 Detailed descriptions of the psychoeducational program are provided by Boise and colleagues (2005) and Schmall and colleagues (2000). PTC is widely used across the United States; however, the existing research on its impact has not directly tested the theoretical framework that guides the intervention. Previous studies on PTC have been detailed descriptive accounts that are helpful for organizations that are interested in offering PTC (Boise et al., 2005; Kuhn, Fulton, & Edelman, 2003). Studies by Boise and colleagues (2005) and Kuhn and colleagues (2003) primarily included caregivers of individuals with AD and tested the intervention by comparing pre- and posttest outcomes, including self-care and health risk behaviors. Self-efficacy was discussed as an outcome but not as an explanatory variable that potentially contributed the change in self-care and health risk behaviors. Another study on PTC also focused on health risk behaviors but specifically considered sociodemographic characteristics of the caregiver and care recipient as its predictors (Won, Fitts, Favaro, Olsen, & Phelan, 2008). The only significant predictor of health risk behaviors was the number of comorbid medical conditions held by the care recipient. SELF-EFFICACY AND SELF-CARE In intervention research, it is important to understand why the intervention resulted in predicted outcomes and to measure the possible active ingredient or explanatory variable. The self-management model on which PTC is based states that self-efficacy is a key factor that influences change in self-care and health risk behaviors. This model is similar to stress process models in the caregiving literature, in that both highlight the role of self-efficacy in predicting self-care and health risk behaviors (Rabinowitz, Mausbac, Thompson, & Gallagher-Thompson, 2007; Yee & Schulz, 2000). Accordingly, the primary objective of the current study is to investigate how PTC affects health risk and self-care behaviors of caregivers. Specifically, the current study aims to examine the role of self-efficacy in reducing the likelihood of health risk behaviors and increasing the likelihood of self-care behaviors (exercise, stress management, relaxation) among family caregivers of persons with AD. The following hypotheses and research questions were tested: Hypothesis 1: There will be significant differences in health risk and self-care behaviors from pre- to post-ptc. Follow-up Research Question 1: Are there differences between spouses and adult children in health risk and self-care behaviors at pre- and post-ptc? Hypothesis 2: Controlling for contextual variables, change in self-efficacy is expected to predict change in health risk and self-care behaviors. Follow-up Research Question 2: Are there differences between spouses and adult children in the extent to which change in self-efficacy predicts change in health risk and self-care behaviors?

152 M. Y. Savundranayagam and M. Brintnall-Peterson METHOD Procedure The study was reviewed and approved by the Institutional Review Board at the University of Wisconsin Extension. Participants were caregivers who signed up voluntarily to participate in the class. The classes were offered by a variety of social service organizations such as local chapters of the Alzheimer s Association and Aging and Disability Resource Centers. The response rate was approximately 85%. Class leaders=facilitators obtained consent from participants prior to the completion of the pre-ptc questionnaire. Caregivers were asked to complete the pre-ptc questionnaire either before or during Session 1. Post-PTC questionnaires were completed either on or after the last PTC session. The post-ptc questionnaire included an open-ended question that asked participants to write about the ways in which PTC helped them with specific concerns or problems related to caregiving. Some of the one-time qualitative responses that were part of the posttest are included in the discussion to augment the quantitative findings. Participants Participants included 325 family caregivers of persons with AD who completed evaluations for PTC classes held in Wisconsin from 2001 to 2004. The average age of caregivers was 65 years, and a little over three-fourths were female. There were almost equal number of wives (38.2%) and daughters-in-law (39.7%) in the caregiving role (see Table 1). An overwhelming majority of participants were White (97%). Approximately 5% of the data was missing. Expectation maximization was used to impute the missing values (Dempster, Laird, & Rubin, 1977; Horton & Kleinman, 2007). Measures: Independent Variables SOCIODEMOGRAPHIC AND BACKGROUND VARIABLES Demographic data was collected pre-ptc and included caregiver age, gender, education, household income, and kinship status (spouse or adult child). Number of sessions attended by each participant was also included in the analyses to control for the effects of exposure to PTC content. Program attendance was strong in the sample; 63% of participants attended all six PTC sessions, 25% attended five sessions, 10% attended four sessions, and 2% attended three or fewer sessions. CAREGIVING DEMANDS The care recipient s level of severity of memory problems was measured using a single-item variable with three levels of mild, moderate, and severe

Self-Efficacy and Self-Care Among Family Caregivers 153 TABLE 1 Description of the Sample (N ¼ 325) Caregiver Average age and range 65 (38 90 years) Gender Male 22.1% Female 77.9% Relationship to person with AD Wife 38.2% Husband 13.8% Daughter 39.7% Son 8.3% Education Grades 0 8 5.2% Grades 9 11 2.5% High school graduate 33.5% Some college 27.1% Graduate coursework 19.7% Household income <$10,000 5.5% $10,000 14,999 10.8% $15,000 20,000 14.8% >$20,000 68.9% Average number of PTC sessions and range 5.5 (2 6) AD ¼ Alzheimer s disease; PTC ¼ Powerful Tools for Caregiving. (Boise et al., 2005; Kuhn et al., 2003). Time spent on personal care, household tasks, and arranging for help were measured using single items that asked participants how often they engaged in the above tasks. The response options included not at all, some days but not every day, and daily or almost daily. CAREGIVER SELF-EFFICACY INTERVENTION OUTCOMES The Caregiver Self-Efficacy Scale is an 11-item measure that was specifically developed for PTC and was based on the skills, behaviors, and attitudes that PTC aimed to influence (Boise et al., 2005). Participants were asked to rate the level of confidence, on a scale of 1 (no confidence) to 5(extremely confident), they had in performing tasks such as getting help with daily tasks, keeping from feeling sad or down in the dumps, and discussing their concerns with family and friends. For example, one item asks, how confident are you that you can discuss openly with the doctor any concerns or problems that you may have related to your caregiving responsibilities. Cronbach s alpha for this measure at pre- and post-ptc was.86 and.87, respectively.

154 M. Y. Savundranayagam and M. Brintnall-Peterson HEALTH RISK BEHAVIORS Health risk was measured by asking participants whether they neglected a particular health-related activity in the past 3 months. Example items included: put off going to the doctor, failed to stay in bed when ill, postponed getting regular checkups or exams, eaten poorly, and put off recreational activities. A score for health risk behavior was calculated by summing the yes responses to the nine items. Cronbach s alpha for this measure at pre- and post-ptc was.74 and.68, respectively. This measure was used previously in studies that evaluated PTC (Boise et al., 2005; Won et al., 2008). SELF-CARE BEHAVIORS The frequency of exercise and stress management techniques used in the past week was measured using single-item questions that asked participants to choose one of the following options: none, <30 minutes, 30 60 minutes, 1 3 hours, and >3 hours (Lorig et al., 1996). Participants were also asked to state the number of times in the past week that they engaged in relaxation activities such as muscle relaxation, prayer, and reading. Analyses Hypothesis 1 (differences in health risk and self-care behaviors from pre- to post-ptc for spouses and adult children) was tested using repeated measures ANOVA and MANOVA. Hypothesis 2 (the role of changes in self-efficacy in predicting changes in health risk and self-care behaviors) was tested using multiple hierarchical regression analysis. Specifically, the regression analysis tested whether changes in self-efficacy predicted changes in health risk and self-care outcomes while controlling for variables related to sociodemographics and caregiving demands. Changes in self-efficacy, health risk, and self-care outcomes were calculated by subtracting pre-ptc scores from post-ptc scores. The first set of variables included sociodemographic factors (age, gender, education, and household income) that may influence access to health promotion opportunities. The second set of variables was associated with caregiving demands (severity of memory problems, and time spent on personal care, household tasks, and arranging for help). The remaining variables were entered separately (e.g., number of sessions attended by each participant, change in self-efficacy, kinship status, and the interaction between kinship status and change in self-efficacy). The following categorical variables were dummy coded prior to regression analyses: gender (0 ¼ male, 1 ¼ female) and kinship status (0 ¼ adult child, 1 ¼ spouse). Finally, change in self-efficacy was mean centered before it was used in the regression analyses to (1) diminish potential multicollinearity between the main effect (of change in self-efficacy) and the interaction (between

TABLE 2 Descriptive Statistics of Major Study Variables and Their Intercorrelations Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1. Age 1.00 2. Gender 0.23 1.00 3. Education 0.34 0.00 1.00 4. Household income 0.13 0.05 0.29 1.00 5. Severity of memory 0.11 0.11 0.10 0.14 1.00 problems 6. Time spent: Personal care 0.14 0.01 0.13 0.20 0.24 1.00 7. Time spent: Household 0.24 0.03 0.14 0.20 0.11 0.45 1.00 tasks 8. Time spent: Arranging 0.03 0.08 0.03 0.14 0.11 0.43 0.46 1.00 help 9. Number of PTC sessions 0.16 0.14 0.15 0.00 0.05 0.01 0.05 0.07 1.00 10. D in Self efficacy 0.07 0.05 0.12 0.01 0.05 0.03 0.03 0.04 0.05 1.00 11. Kinship status 0.77 0.11 0.30 0.07 0.14 0.16 0.27 0.09 0.09 0.00 1.00 12. D in health risk 0.06 0.05 0.00 0.08 0.09 0.07 0.02 0.01 0.00 0.32 0.02 1.00 behaviors 13. D in exercise 0.03 0.02 0.05 0.01 0.04 0.03 0.00 0.07 0.08 0.14 0.01 0.19 1.00 14. D in stress management 0.09 0.11 0.02 0.08 0.08 0.00 0.02 0.08 0.14 0.24 0.07 0.12 0.13 1.00 15. D in # of relaxation 0.11 0.15 0.00 0.10 0.01 0.03 0.07 0.03 0.10 0.20 0.02 0.02 0.02 0.14 1.00 activities M 64.96 3.90 3.47 2.06 1.80 2.50 2.16 5.48 0.55 1.07 0.37 0.57 1.21 SD 11.85 1.26 0.89 0.67 0.82 0.66 0.68 0.78 0.58 2.04 1.25 1.37 6.44 Note: PTC ¼ Powerful Tools for Caregiving. Correlation is significant at p <.05 level. 155

156 M. Y. Savundranayagam and M. Brintnall-Peterson change in self-efficacy and kinship status) and (2) to make parameter estimates easier to interpret (Aiken & West, 1991). Table 2 includes descriptive statistics and correlations of variables used in the analyses. RESULTS Differences in Health Risk and Self-Care Behaviors From Pre- to Post-PTC Changes in participants health risk and self-care behaviors by kinship status over the course of PTC were examined using a two-factor repeated measures ANOVA for health risk behaviors and number of relaxation activities, and repeated measures MANOVA for time spent exercising and time spent on stress management techniques. There were no significant interactions between kinship status and time for any of the outcomes. The main effects for kinship status, F(1,323) ¼ 13.70, p <.001, gp 2 ¼.04, and time, F(1,323) ¼ 90.08, p <.001, gp 2 ¼.22, were significant for health risk behaviors. There were significant reductions in health risk behaviors from preto post-ptc for spouses and adult children. Adult children also engaged in a greater number of health risk behaviors compared to spouses. There was a significant time main effect for number of relaxation activities, F(1,323) ¼ 11.51, p <.01, gp 2 ¼.03. Over the course of PTC, spouses and adult children engaged in more relaxation activities. Finally, there was a significant main effect for time on exercise and stress management techniques, k(2, 322) ¼ 38.44, p <.001, gp 2 ¼.19. Follow-up univariate tests revealed that both groups spent more time on exercise, F(1,323) ¼ 28.90, p <.001, gp 2 ¼.08, and stress management techniques, F(1,323) ¼ 57.16, p <.001, gp 2 ¼.15, from pre- to post-ptc. Repeated measures ANOVA was used to test whether there was a kinship status by time interaction for self-efficacy. Results revealed main effects for kinship status, F(1,323) ¼ 9.66, p <.01, gp 2 ¼.03, and time, F(1,323) ¼ 282.27, p <.001, gp 2 ¼.47, but no interactions. Although adult children scored higher on self-efficacy over the course of PTC, both groups showed a significant increase in self-efficacy from pre- to post-ptc. Changes in Self-Efficacy as a Predictor of Changes in Health Risk and Self-Care Behaviors The separate sets of regression analyses revealed that there were no significant group differences in the extent to which change in self-efficacy predicted all four outcomes (health risk behaviors, exercise, stress management techniques, and relaxation activities). Moreover, the full model for change in exercise was not significant, F(12, 324) ¼ 1.08, p ¼.38. As a result, follow-up analyses were conducted with three sets of regression analyses that did not include the kinship status and interaction variables (see Table 3). The full

TABLE 3 Hierarchical Regression Analyses for Variables Predicting Change in Health Risk and Self-Care Outcomes Change in health risk Change in time spent on Change in the number of behaviors a stress management b relaxation activities c Predictor variable B SEB b B SEB b B SEB b Step 1 Age 0.00 0.01 0.02 0.01 0.01 0.09 0.06 0.03 0.11 Gender 0.10 0.27 0.02 0.33 0.18 0.10 2.10 0.87 0.14 Education 0.09 0.10 0.06 0.04 0.07 0.04 0.25 0.31 0.05 Household income 0.00 0.01 0.02 0.12 0.09 0.08 0.73 0.41 0.10 Step 2 Severity of memory problems 0.21 0.17 0.07 0.20 0.11 0.10 0.20 0.54 0.02 Time spent: Personal care 0.25 0.16 0.10 0.11 0.11 0.06 0.13 0.50 0.02 Time spent: Household tasks 0.18 0.20 0.06 0.05 0.13 0.02 1.09 0.63 0.11 Time spent: Arranging help 0.09 0.19 0.03 0.17 0.13 0.09 0.17 0.60 0.02 Step 3 Number of PTC sessions 0.02 0.14 0.01 0.28 0.10 0.16 1.17 0.46 0.14 Step 4 Change in self-efficacy 1.11 0.19 0.32 0.54 0.13 0.23 2.07 0.60 0.19 Note: PTC ¼ Powerful Tools for Caregiving. Coefficients are from the Step 4 model. PTC ¼. a For the regression on change in health risk behaviors, R 2 ¼.01 for Step 1 (p ¼.42); DR 2 ¼.02 for Step 2 (p ¼.23); DR 2 ¼.00 for Step 3 (p ¼.86); DR 2 ¼.10 for Step 4 (p <.001). b For the regression on change in time spent on stress-management, R 2 ¼.02 for Step 1 (p ¼.10); DR 2 ¼.01 for Step 2 (p ¼.30); DR 2 ¼.03 for Step 3 (p <.01); DR 2 ¼.05 for Step 4 (p <.001). c For the regression on change in the number of relaxation activities, R 2 ¼.04 for Step 1 (p <.05); DR 2 ¼.00 for Step 2 (p ¼.68); DR 2 ¼.02 for Step 3 (p <.01); DR 2 ¼.03 for Step 4 (p <.01). p <.05. 157

158 M. Y. Savundranayagam and M. Brintnall-Peterson model for change in health risk behaviors was significant, F(10, 324) ¼ 4.59, p <.001, and explained 13% of the variance. Results indicated change in self-efficacy was the only significant predictor, accounting for 10% of the variance in changes in health risk behaviors. Specifically, the increase in self-efficacy predicted the decrease in health risk behaviors (b ¼.32, p <.001). The full model for change in the frequency of using stress management techniques was significant, F(10, 324) ¼ 4.10, p <.001, and explained 12% of the variance. Results indicated the number of PTC sessions attended and change in self-efficacy uniquely accounted for 3% and 5%, respectively, of the variance in changes in frequency of stress management techniques. Specifically, the more PTC sessions attended (b ¼.16, p ¼.004) and increases in self-efficacy (b ¼.23, p <.001) predicted the increase in use of stress management techniques. Finally, the full model for change in the number of relaxation activities was significant, F(10, 324) ¼ 3.54, p <.001, and explained 10% of the variance. Results indicated that gender, number of PTC sessions attended, and change in self-efficacy uniquely accounted for 2%, 2%, and 3%, respectively, of the variance in changes in the use of relaxation activities. Specifically, females (b ¼.14, p ¼.017) were more likely to the increase their use of relaxation activities. The more PTC sessions attended (b ¼.14, p ¼.011) and increases in self-efficacy (b ¼.19, p ¼.001) predicted the increase in use of relaxation activities. DISCUSSION Overall, the findings from the current study are similar to previous studies on PTC (Boise et al., 2005; Kuhn et al., 2003; Won et al., 2008) in that they illustrate the positive effect of PTC on health risk and self-care behaviors such as relaxation, exercise, and stress management. Moreover, participants showed improvements in self-efficacy over the course of PTC (e.g., My confidence level has improved each week ). The following comments reflect the newfound awareness of self-neglect among participants that prompted self-care: The emphasis on self-care was critical for me as this neglect of my needs was moving me into early burnout. It helped me to see I have to take care of myself so I can be around to take care of my husband in the future. Participation in PTC gave caregivers permission to engage in self-care behaviors, as illustrated by the following comments: Validated my need to care for myself, go to work, meet friends for coffee, lunch, etc. ; Gave myself permission to thrive, not just survive, and to take better care of myself. These findings clearly illustrate that, though family members have been actively providing for their relative with AD, their own needs have been either neglected or not affirmed in the past. Participating in PTC brought their unmet needs to the forefront and, more importantly, acknowledged the value of resolving unmet needs.

Self-Efficacy and Self-Care Among Family Caregivers 159 What is distinct about the current study is that it also examined the extent to which changes in self-efficacy explained the favorable changes in health risk and self-care outcomes. Spouses and adult children did not differ in terms of the extent to which change in self-efficacy explained changes in health risk and self-care outcomes. Although change in self-efficacy did not explain the improvements in time spent on exercise, it was most influential in explaining the reduction in health risk behaviors, followed by increases in time spent on stress management and increases in relaxation activities. Statements such as PTC gave me confidence to take care of self and gave me confidence in myself so that I can better handle situations that arise illustrate how self-efficacy influenced self-care. Getting family members to engage in self-care is challenging because the individual with the chronic illness such as AD is the focus of social and health care systems. To foster family-centered care, it is all the more important that psychoeducational programs like PTC actively focus on boosting self-efficacy among family caregivers so that they can also advocate for their own care. The findings revealed that the number of sessions attended uniquely contributed to the increases in time spent on stress management and increases in the number of relaxation activities. Each PTC session included developing action plans for the upcoming week and reporting on action plans from the previous week. Action plans involved creating opportunities for the caregiver to engage in self-care activities. Participants were also encouraged to develop action plans around ways they could take better care of themselves or reduce stress. Therefore, the more sessions attended, the more practice participants had creating and following through on action plans that involved stress management or relaxation (e.g., action plans helped me plan how to get some time to myself ). Moreover, participants received weekly feedback and encouragement on action plans. This is critical in the case of caregivers of persons with AD because the PTC group might be their only opportunity to receive positive and affirming feedback from others. Moreover, by having caregivers report on their action plans each week, their self-efficacy increased their ability to develop action plans that could be accomplished successfully. This is evidenced by the following comments: Motivated for movement on previously planned action and I found I can do better planning and enjoy my role more with goals and thinking before I act. In addition to action plans, each PTC session featured a different relaxation technique such as guided imagery, muscle relaxation, and deep breathing. Relaxation tapes were also available for participants to borrow or buy. Some of the participants took advantage of this opportunity and used the relaxation methods daily; those who attended all of the classes were taught a variety of relaxation techniques to help maintain self-care. Limitations Although the sample size was relatively large for an intervention study, there are several limitations of the current study in terms of the sample. First, it

160 M. Y. Savundranayagam and M. Brintnall-Peterson included a convenience sample of individuals who self-selected to take PTC and complete the pre- and post-ptc questionnaires. Second, the sample was relatively homogeneous: most participants were female and White. However, this ethnic composition in the current study reflects the less urban areas of Wisconsin where PTC sessions were offered. Nonetheless, outreach to caregivers of color is critical and an area that is worthy of more research. Finally, there was no control group, and as a result one cannot be certain that the positive outcomes are due to the intervention or indicative of general change and adaptation over time. Unlike previous work on PTC (Boise et al., 2005), the current study reports on the immediate impact of PTC on caregivers. Longer term follow-up will be helpful not only to evaluate maintenance effects but also to examine whether there are delayed intervention effects for participants who did not show immediate improvements in self-efficacy, or health risk and self-care behaviors. Although increased levels of self-efficacy improved the likelihood that caregivers engaged in self-care behaviors, self-efficacy only explained part of the variance in the outcome measures. Future research should consider other mechanisms (or mediators) that explain improvements in self-care and health risk behaviors. Moreover, other participant characteristics (or moderators) apart from kinship status might influence the degree of improvement in self-care outcomes. CONCLUSION Testing theoretical models that underlie interventions by investigating the mechanisms that contribute to positive health outcomes for caregivers is increasingly important. The self-management (Lorig & Holman, 2003; Lorig et al., 1996) and stress process models (Yee & Schulz, 2000) hypothesize that self-efficacy plays a pivotal role in maintaining caregiver health. Previous research findings indicated that lower levels of self-efficacy are linked with negative health behaviors among caregivers of persons with AD (Rabinowitz et al., 2007). As such, the current study directly tested whether changes in self-efficacy over the course of PTC predicted changes in health risk and self-care behaviors. The findings have important implications for psychoeducational interventions in general and for future PTC classes. First, the findings illustrate that participants who increased their level of self-efficacy were able to reduce health risk behaviors and increase their involvement in self-care behaviors. Therefore, it is important for class leaders to be attuned to self-efficacy levels of their participants as they progress through the program. Perhaps PTC should be targeted to caregivers who have low levels of self-efficacy at the start of the program. Second, trained class leaders need to know that attending all six PTC sessions increases the likelihood that caregivers will engage in self-care behaviors. Potential PTC participants need to

Self-Efficacy and Self-Care Among Family Caregivers 161 be encouraged to stay the course to achieve the full self-care benefits of the program. This might be accomplished in a variety of ways, including offering respite when PTC sessions are held. Some organizations that offer PTC do provide respite and they should be commended for integrating services so that family caregivers can learn more about self-care. Overall, the study s findings highlight the importance of creating awareness about the prevalence of self-neglect among family caregivers as a necessary first step in promoting family-centered health care practice. We are finally reaching a point where service providers and health professionals are realizing that caregiver health is important. It is up to professionals to use psychoeducational programs to inform family caregivers about the deleterious effects of their self-neglect and to empower them with tools to engage in self-care behaviors. Such acknowledgment and support might help family members be better able to maintain their roles as caregivers without risking their own health. REFERENCES Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Bédard, M., Pedlar, D., Martin, N. J., Malott, O., & Stones, M. J. (2000). Burden in caregivers of cognitively impaired older adults living in the community: Methodological issues and determinants. International Psychogeriatrics, 12, 307 332. Boise, L., Congleton, L., & Shannon, K. (2005). Empowering family caregivers: The powerful tools for caregiving program. Educational Gerontology, 31, 1 14. Clipp, E. C., & George, L. K. (1993). Dementia and cancer: A comparison of spouse caregivers. The Gerontologist, 33, 534 541. Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977). Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, B, 39, 1 38. Gallant, M. P., & Connell, C. M. (1997). Predictors of decreased self-care among spouse caregivers of older adults with dementing illnesses. Journal of Aging and Health, 9, 373 395. Hooker, K., Monahan, D. J., Bowman, S. R., Frazier, L. D., & Shifren, K. (1998). Personality counts for a lot: Predictors of mental and physical health of spouse caregivers in two disease groups. Journal of Gerontology, 53B, 73 85. Horton, N. J., & Kleinman, K. P. (2007). Much ado about nothing: A comparison of missing data methods and software to fit incomplete data regression models. American Statistician, 61(1), 81 90. Kuhn, D., Fulton, B., & Edelman, P. (2003). Powerful tools for caregivers: Improving self-care and self-efficacy of family caregivers. Alzheimer s Care Quarterly, 4, 189 201. Lorig, K. R., & Holman, H. (2003). Self-management education: History, definition, outcomes, and mechanisms. Annals of Behavioral Medicine, 26, 1 7.

162 M. Y. Savundranayagam and M. Brintnall-Peterson Lorig, K. R., Stewart, A., Ritter, P., Gonzalez, V., Laurent, D., & Lynch, J. (1996). Outcome measures for health education and other health care interventions. Thousand Oaks, CA: Sage. Ory, M. G., Hoffman, R. R., Yee, J. L., Tennstedt, S., & Schulz, R. (1999). Prevalence and impact of caregiving: A detailed comparison between dementia and nondementia caregivers. The Gerontologist, 39(2), 177 185. Pinquart, M., & Sorensen, S. (2003). Associations of stressors and uplifts of caregiving with caregiver burden and depressive mood: A meta-analysis. Journal of Gerontology, 58(B), P112 P128. Rabinowitz, Y., Mausbac, B., Thompson, L., & Gallagher-Thompson, D. (2007). The relationship between self-efficacy and cumulative health risk associated with health behavior patterns in female caregivers of elderly relatives with Alzheimer s dementia. Journal of Aging & Health, 19(6), 946 964. Savundranayagam, M. Y., Hummert, M. L., & Montgomery, R. J. V. (2005). Investigating the effects of communication problems on caregiver burden. Journal of Gerontology, 60B(1), S48 S55. Schmall, V. L., Cleland, M., & Sturdevant, M. (2000). The caregiver help book (use in the powerful tools for caregiving). Portland, OR: Oregon Gerontological Association, Legacy Health System. Schulz, R., & Beach, S. R. (1999). Caregiving as a risk factor for mortality: The caregiver health effects study. Journal of the American Medical Association, 282, 2215 2219. Vitaliano, P. P., Scanlan, J. M., Zhang, J., Savage, M. V., Hirsch, I., & Siegler, I. C. (2002). A path model of chronic stress, the metabolic syndrome, and coronary heart disease. Psychosomatic Medicine, 64, 418 435. Vitaliano, P. P., Zhang, J., & Scanlan, J. M. (2003). Is caregiving hazardous to one s physical health? A meta-analysis. Psychology Bulletin, 129, 946 972. Won, C. W., Fitts, S. S., Favaro, S., Olsen, P., & Phelan, E. A. (2008). Communitybased powerful tools intervention enhances health of caregivers. Archives of Gerontology and Geriatrics, 46(1), 89 100. Yee, J. L., & Schulz, R. (2000). Gender differences in psychiatric morbidity among family caregivers: A review and analysis. The Gerontologist, 40, 147 164.