Public Health Nursing Vol. 9 No. 2, pp. 103-108 0737-1 209/92/$6.00 0 1992 Blackwell Scientific Publications, Inc.
Description of Learned Response to Chronic Illness: Depressed versus Nondepressed Self-Help Class Participants Carrie Jo Braden, Ph.D., R.N.
Abstract The purpose of this study was to describe the effectiveness of a community-based health-promotion program that was collaboratively designed and that is currently being implemented by lay persons who have the diagnosis of systemic lupus erythematosus (SLE) and by health professionals. Two groups of SLE self-help course participants contributed data for this theory-guided preexperimental program impact study. Braden’s self-help model provided the theoretical framework. The variables addressed were severity of illness, limitations, uncertainty, enabling skill self-efficacy, self-worth, and life quality. The groups consisted of 35 subjects having low depression scores ( ~ 7 7 and ) 37 with high depression scores (a272) on a measure having a possible range of scores from 0 to 400. Data were collected at three points: at the beginning of class 1, after class 7 (7 wks later), and two months after completion of the course. A group (2) by time (3) by measures (10) MANOVA analysis procedure was used with a follow up univariate F test and Newman-Keuls multiple comparison procedure. Significant change was evident for the level of depression group and for three interaction effects of group X time, group X measure, and group X time X measure.
The purpose of this study was to describe the effectiveness of a community-based health-promotion program that was designed and implemented with persons who have the diagnosis of systemic lupus erythematosus (SLE). The description of program effectiveness addresses the learned response of two groups of participants in an SLE self-help course (Braden et al., 1987). One group represents the upper end of an adaptation continuum based on low depression scores and the other represents the lower end of the continuum based on high depression scores. Learned response variables included severity of illness, limitations, uncertainty, enabling skills, self-efficacy, self-worth, and life quality. Health promotion for persons with disabilities has received little notice, perhaps due to lack of focused attention on integrating polar views about system versus individual approaches, as well as lack of a framework for understanding the dynamics of how disabled persons engage in health-promoting strategies. Little research has been directed to health care attitudes and behaviors of persons with disabilities that could lead to interventions enhancing their health-promoting behaviors (Becker et al., 1989). Yet, if such behaviors are important for the nonimpaired population, they are more critical for those whose conditions leave them with a thinner margin of health (Becker et al., 1987; DeJong & Hughes, 1982). BACKGROUND
Several recent definitions provide a basis for viewing health promotion as an optimum vehicle for improving the health of persons with disabilities. Goodstadt, SimpResearch supported b y grant NCNR NRO 1686, from the Nason, and Loranger (1987), Pender (1987), and Green and tional Arthritis Foundation. Address correspondence to Carrie Jo Braden, University of Ar- Raeburn (1990) provided definitions from a perspective that places health professionals in the role of consultant, izona, College of Nursing, Tucson, A Z 85721.
103
104 Public Health Nursing
Volume 9 Number 2 June 1992
hypotheses from three alternative learning theories and incorporates the supported specifications from learned resourcefulness theory (Rosenbaum, 1988), learned helplessness theory (Seligman, 1975), and instrumental passivity theory (Baltes, 1982). The concepts addressed by this study of persons who participated in a health education program designed to enhance a learned self-help response to SLE are based on Braden’s self-help model construct specifications for 1) the stimulus for learning, 2) the negative reinforcers or adversities inherent in chronic illness, 3) the discriminant stimuli that act as “on” or “off” switches for involvement in learning, and 4) the outcome behaviors representing positive goals and serving as positive reinforcers for further learning. Robinson (1 982) defined a learned self-help response to negative life events as an informed process of facing definable, manageable adversities by maintaining control of everyday problems. The opposite response, succumbing, was described by Wright (1960), and addressed as learned helplessness by Seligman (1975) and as instrumental passivity by Baltes (1982). It can be defined as an uninformed process of withdrawing from definable, manageable adversities by giving in to everyday problems. Although it is known that most persons with a chronic illness do not give up and, in fact, learn by trial and error how to manage everyday problems themselves (Haug & Lavin, 1983), more remains to be known about how this learned self-help response works to turn the uncertainties and limitations of a disease like SLE into definable, manageable adversities. The self-help model poses that the perceived severity of the chronic illness initiates learning about living with it. The increased exposure to related adversities opens the door to a new learning frame. The model also hypothesizes that a repertory of enabling skills can serve as a cue to go ahead with discovering what to do on a day-to-day basis to reach desired goals. Some learning theorists, specifically Bandura (1977), assert that one’s THEORETICAL MODEL level of self-efficacy also mediates aversive aspects of A variety of theoretical models are also being used to illness. describe the dynamics of how persons engage in healthIt is also likely that negative mediators, for example, promoting behaviors despite chronic debilitating dis- depression, reduce involvement in learning. Depression ease. Braden’s self-help model is based on the assump- snuffs out a continued commitment to trying to overtions that learning is a fundamental process underlying come problems (Seligman, 1975). Maintenance of selfadaptation, that chronic illness represents exposure to worth and life quality are often goals targeted by pera new learning environment, and that learning poten- sons who are diagnosed with a chronic illness. Findings tially occurs within specific contexts involving the in- of Becker et al. (1989) indicated that more than half the teraction of intrapersonal, interpersonal, and environ- sample with disabilities had health-related goals that mental factors. The model has been used to describe focused on functioning well. However, not all perlearned self-help and learned helplessness response sons have a sufficient level of enabling skills or of selfamong persons with the diagnosis of rheumatoid arthri- efficacy to cue continued involvement in learning in tis or arthritis-related conditions (Braden, 1990). The order to reach desired goals, particularly in the face of self-help model emerged from the competitive testing of depression.
advocate, mediator, and supporter, and from an individual perspective that invests lay persons with decision power. These definitions, focusing on actualizing health potentials rather than on the eradicating evidence of disease or infirmity, concur that individuals diagnosed with chronic illness or living with a disability can be healthy. Emphasis is placed on self-care and mutual aid prerogatives, and provides for shared responsibility for health between individuals and systems (Green, 1986). A descriptive study of health-promoting behaviors among persons with disability demonstrated that many adults place great emphasis on the functional, adaptive, and self-actualizing aspects of health (Becker et al., 1989). Nearly three-fourths of the sample in this study perceived themselves as being healthy. Pender’s model of health promotion was applicable to persons with disabilities. A combination of attitudinal factors, such as self-efficacy, and demographic factors, such as age and sex, were related to self-reports of engaging in healthpromoting activities such as exercise, nutrition, and stress management (Becker et al., 1989). This study reports the results of a program that was based on the definition of health promotion as being the implementation of educational, organizational, economic, and environmental supports to maintain and enhance health (Goodstadt et al., 1987; Green & Raeburn, 1990; Pender, 1987). The program evaluation, a preexperimental design advocated by Chen (1990) as appropriate when the treatment (self-help course) directly acts on characteristics of the problem (a desire to achieve health), also incorporated lay persons in the evaluation planning process. Chen (1990) proposed theory-guided strategies for impact evaluation using the one group pretest-posttest design when costs of an experimental design are prohibitive. The preexperimental design provides worthwhile information about a program’s impact.
Braden: Enabling Skill, Self-care 105
DESCRIPTION OF THE INTERVENTION
tional Office in Atlanta, Georgia, and offered at 17 sites The self-help course incorporated the beliefs, needs, across the country. Class sites included four in Califorand concerns of persons with SLE by relying on nia and Texas, two in New Jersey and Kansas, and one consumer participation in developing the curriculum, each in Colorado, Oregon, Georgia, Iowa, and Illinois. rather than only on health professionals. The course Subjects were recruited into the study over four years uses lay teachers in partnership with health professional from early 1987 to late 1990. For attenders willing to participate in the study, data were collected at three teachers. points: during the first class session, after the last class The course activities focus on problem solving, cog(7 wks later), and two months after completing the nitive reframing, and enhancing belief in self. Exercise class. The Arthritis Foundation arranged for data coland relaxation activities were incorporated as methods of strengthening self-efficacy relative to specific health lection and mailing of completed forms to the invesbehavior. All activities are designed to enhance partic- tigator for data entry and analysis. Data on number of classes attended was available on 104 subjects. ipants’ interaction. The process of sharing with others The number ranged from one to seven classes attended is facilitated as a means of building self-worth as well as (mean 5.6 classes, SD 1.5). of adding to a repertory of problem-solving skills and to a set of alternative solutions for common problems. Demographics While trial-and-error over time may effectively strengthen enabling skills and self-efficacy and reduce depression Ninety-six percent (263) of the sample were women. in some persons, in others learning slows or even stops. Ages ranged from 21 to 83 years (mean 46 yrs). Most The goals of the self-help course were to restart trial- subjects were married (70%). Ethnic composition was and-error learning for participants who had slowed or 84% Caucasian, 7% black, 1.5% Asian, 0.7% American stopped involvement, and to maintain a self-help re- Indian, 2.9% Mexican, 0.4% Puerto Rican, 2.2% other, sponse in those who had continued to be involved in with 1% missing data. Fifty-nine percent had some collearning. Activities were designed to reduce a percep- lege education; only 7.4% had less than a high school tion of dependency, to facilitate uncertainty manage- diploma, with the lowest level of education being sevment, to strengthen enabling skills and self-efficacy, and enth grade. Forty-one percent worked outside the to build self-worth and subsequent improved life qual- home, with 30% of these indicating they were full-time ity. The theoretical basis for describing the learned employees. A nearly equal number (29%) indicated they response of participants specified the concepts of were full-time homemakers. Ten percent were retired perceived illness severity (stimulus), limitations and un- and 3% were seeking work. Fifteen percent reported certainty (negative reinforcers), enabling skill and self- they were disabled. The number of years ill ranged from efficacy (discriminant stimuli), and self-worth and life newly diagnosed to 30 years (mean 6.7 yrs). Twentyquality (desired outcomes). Class-tied variables in- four percent of the sample had been ill one year or cluded knowledge about SLE and activities to facilitate less. pain management, energy conservation, and mainteMeasures nance of functional status. Thus, the question addressed by this study was; what changes occur over time in per- The questionnaire was developed by health professionceived illness severity, limitations, uncertainty, enabling als with the help of persons having diagnosis of SLE skills, self-efficacy, self-worth, life quality, SLE knowl- who participated in establishing the original self-help edge, and the range and number of rest, relaxation, and course. The purpose was to put in place a procedure for exercise activities between participants who enter the collecting descriptive information about participants at class with high depression versus those who enter with the time the course was being distributed nationally. low depression? The developers met with several other persons, includMETHODS Procedure and Sample Human subjects criteria were met and approved by a human subjects review committee from a College of Nursing located in southwestern United States. A convenience sample of 273 subjects with diagnosis of SLE resulted from the recruitment of participants in a selfhelp course sponsored by the Arthritis Foundation Na-
ing this investigator, to collect instruments that had been used to measure various aspects of chronic illness. They then independently selected the items from the instruments they thought would be appropriate, and also created some of their own. The following variable definitions are consistent with variables and definitions provided in Braden’s (1990) self-help model. Severity of illness, one’s perceived level of infliction due to disease course characteristics, was measured by three visual analog (VAS) formatted items requesting in-
106 Public Health Nursing
Volume 9 Number 2 June 1992
formation about common SLE symptoms that had occurred during the past month. The 10-mm, horizontal VAS response line for the first item requesting response about muscle/joint pain had end point anchors of “as bad as pain could possibly be” and “no pain.” The end points for the second item response line were “no fatigue” and “as bad as fatigue could possibly be.” The third item regarding the amount of muscle/joint stiffness was anchored by “stiffness as bad as could be” and “no stiffness. High scores indicated greater symptom severity. The time 1 scores for the three summed items ranged from 0 to 281 (mean 157.5, SD 65, skewness -0.4), time 2 scores from 0 to 287 (mean 147.I , SD 60.2, - skewness 0.1), and times 3 from 0 to 285 (mean 143,1, SD 62, skewness 0.2). Internal consistency reliability estimates using Cronbach’s standardized alpha coefficient were 0.74, 0.81, and 0.79 for time 1, 2, and 3. Limitations, perceived level of inability to do things for oneself, was measured by a single item. The item asked respondents to select one of four statements that best describes what they are able to do. The four statements were “I can do everything I want to do”; “I can do most of the things I want to do, but have some limitations”; “I can do some, but not all, of the things I want to do”; and “I have many limitations and I can do hardly any of the things I want to do.” The item was scored with the highest score, 4, indicating greatest limitation. At time 1 the mean was 2.3 (SD 0.7, skewness 0.2), at time 2 it was 2.2 (SD 0.6, skewness 0.5), and at time 3 it was 2.2 (SD 0.6, skewness 0.2). Uncertainty, one’s perceived amount of ambiguity about treatment effectiveness, was measured by three items, each having a VAS response line with end points of “not at all certain” and “very certain.” The items concerned certainty in ability to use self-care techniques, effectiveness of medications to improve SLE, and ability to decrease pain or stiffness. High scores indicated low levels of certainty. Summed scores at time 1 ranged from 0 to 282 (mean 128.6, SD 66.5, skewness 0.2), at time 2 from 1 to 300 (mean 96.9, SD 62.7, skewness 1 . 1 ) , and at time 3 from 0 to 300 (mean 98.7, SD 66.7, skewness 0.8). The standardized Cronbach’s alpha for times 1 , 2, and 3 were 0.66, 0.79, and 0.79. Depression, one’s level of despondency, was measured by four items having a VAS response format. The response line for the first item, “how have you been feeling in general?”, was anchored with “in excellent spirits” and “in very low spirits.” The response line for the second item, “have you felt so sad, discouraged, hopeless, or had so many problems that you wondered if anything was worthwhile?” was anchored by “extremely so, to the point I have just about given up” and “not at all.” The response line for the third item, “have you felt down hearted and blue?”, was anchored by “all ”
of the time” and “none of the time.” The last item asking, “how depressed have you been?” had response line end points of “very depressed” and “not at all depressed.” The items were scored in the direction of depression. Time 1 depression scores ranged from 0 to 381 (mean 174.6, SD 97.8, skewness 0.2), time 2 scores from 0 to 400 (mean 147.5, SD 92.7, skewness 0.6), and time 3 from 0 to 396 (mean 153.2, SD 97.5, skewness 0.4). Internal consistency estimated by the standardized Cronbach’s alpha was 0.90 for time 1 , 0.86 for time 2, and 0.92 for time 3. Enabling skill, one’s perceived level of ability to manage adversity, was measured by six items drawn from Rosenbaum’s (1980) 36-item self-control schedule. A VAS response format with end points of “true about me” and “not true about me” was used. Items addressed cognitive reframing skills: “often by changing my way of thinking, I am able to change my feelings about almost anything”; “when I am depressed I try to keep myself busy with things 1 like to do”; and “when I feel pain, I try to divert my thoughts from it.” Items also addressed problem-solving skills: “when faced with a difficult problem, I try to approach its solution in a systematic way”; and “I keep track of how well a treatment works for me and I usually plan my work when I’m faced with a lot of things to do.” The items were scored in a positive direction for high-enabling skill. At time 1 the summed score on the six items ranged from 0 to 600 (mean 393.3, SD 129.1, skewness -0.9), at time 2 from 33 to 600 (mean 416.3, SD 117, skewness - 1 . I ) , and at time 3 from 0 to 600 (mean 43 -, SD 116.7, skewness - 1.2). The standardized Cronbach’s alpha at time 1 was 0.82, at time 2 0.85, and at time 3 0.87. Self-efficacy, the strength of conviction that one can do what is necessary to control a primary SLE-related symptom, was measured by a single item having a VAS response format. The response line for the item, “how satisfied are you with your ability to control fatigue?” was anchored by “very dissatisfied” and “very satisfied.” Items were scored in a positive direction for greater self-efficacy. Scores on the item ranged from 0 to 100 (mean 39.6, SD 28.3, skewness 0.5) for time 1 , from 0 to 100 (mean 52.8, SD 24.8, skewness-0.2), at time 2, and from 0 to 100 (mean 56.3, SD 25.3, skewness -0.4) at time 3. Self-worth, level of positive feelings about one’s own being, was measured by 10 items having a VAS response format with end points of “true about me” and “not true about me.” Items concerned ways subjects felt about themselves in general. They were scored in a positive direction for self-worth. Scores at time 1 ranged from 0 to 1000 (mean 629.9, SD 223.6, skewness-0.4), at time 2 from 5 to 1000 (mean 682.8, SD 217.4, skewness - l . l ) , and at time 3 from 0 to 1000 (mean 667.4,
Braden: Enabling Skill, Self-care 107
SD 232.7, skewness -0.8). The standardized Cronbach’s alpha at time 1 was 0.90; at time 2 was 0.92, and at time 3 was 0.93. Life quality, perceived level of satisfaction with one’s personal situation, was measured by three items drawn from Campbell, Converse, and Rogers’s (1976) 10-item semantic differential scale, inventory of well-being. The items selected were “my present life is enjoyable-miserable,” or “full-empty,” and “in thinking about my life as a whole, I am completely satisfied-completely dissatisfied.” Higher scores reflected greater life quality. Responses on the three summed items at time 1 ranged from 0 to 300 (mean 173.4, SD 67.1, skewness -0.3), at time 2 from 0 to 300 (mean 187.4, SD 63.8, skewness - 0.4), and at time 3 from 2 to 300 (mean 184.6, SD 68.6, skewness - 0.4). The standardized Cronbach’s alpha coefficient for time 1 was 0.85, for time 2 was 0.85, and for time 3 was 0.90. Three additional variables, SLE knowledge, activity number, and activity range, were included to address information gained through class content and through rest, relaxation, heat, and exercise alternatives that were incorporated into class activities for pain management and for strengthening self-efficacy. Knowledge about SLE, the level of information about disease characteristics and treatment alternatives, was measured by seven multiple-choice items. The range of correct answers at time 1 was 0 to 7 (mean 4.7, SD 1.5, skewness -0.6), at time 2 from 1 to 7 (mean 5.5, SD 1.2, skewness -0.7), and at time 3 from 1 to 7 (mean 5.5, SD 1.2, skewness - 0.9). Internal consistency reliability as estimated by Cronbach’s alpha was 0.47 at time 1 , 0.45 at time 2, and 0.54 at time 3. Rest, relaxation, heat exercise range, the number of different types of pain-management and self-strengthening activities engaged in during one week, was measured by counting the number of different types of activities reported from a check list. Subjects also listed additional physical activities engaged in during a week’s time. These were categorized, and the number of different activity categories added to the reported number of activities from the provided list. The range of activities engaged in at time 1 was 0 to 10 (mean 3.3, SD 1.8, skewness O S ) , at time 2 0 to 10 (mean 3.9, SD 1.9, skewness - 0. l), and at time 3 0 to 1 1 (mean 4.2, SD 1.9, skewness 0.8). Rest, relaxation, heat, exercise number, the total number of times pain-management and self-strengthening activities were engaged in during a week, was measured by summing the reported number of times for each activity. At time 1 the number of times activities were engaged in ranged from 0 to 91 (mean 13.6, SD 11.2, skewness 2) and at time 3 from 0 to 70 (mean 16.3, SD 10.8, skewness 1.5).
RESULTS A group (2) time (3) by measures (10) MANOVA procedures was used to answer the question about the pattern of change in learned response variables and in class-tied variables occurring over time in two groups, those having high depression scores at time of class entry and those having low depression scores. The MANOVA procedure was selected because of the desire to investigate a variable, depression, that could contribute most to separation of learned response patterns (Bray & Maxwell, 1989, p. 11). The depression groups were formed from subjects having depression scores above or equal to a + 1 SD (37 subjects with mean scores 2272) and those with scores below or equal to a - 1 SD (35 subjects with mean scores 677). The results are presented in Table 1. Significant changes were found between high depression and low depression as a main effect. Three significant interaction effects were found for depression group and time, measure, and time and measure. The univariate F test and Newman-Keuls multiple comparison procedure using P =s0.05 were selected to describe patterns of change within each group. These procedures revealed seven sources of significant changes in the high depressed group and five in the low depressed group. No significant changes were found for either group over time for severity of illness or limitation. Uncertainty changed significantly only for the high depression group, with a reduction in uncertainty occurring between times 1 and 2 and times 1 and 3. Both groups changed significantly over time, with enabling skill, self-efficacy, and self-worth increasing between times 1 and 2 and times 1 and 3 for the high depression group. The low depression group evidenced significantly more enabling skill, self-efficacy, and self-worth only from time 1 to time 3. Life quality increased significantly over time only for those in the high depression group. The comparisons for SLE knowledge have to be viewed with caution, given the low estimates for interTABLE 1. Multivariate Analysis of Variance of Depression Group by Time and Measure
Source of Variation
df
F
Group Time
1170 21140 91630 21140 91630 1811264 1811 264
8.3* 0.9 0.3 30.lt 20.2t 0.3 8.9t
Measure Group x time Group x measure Time x measure Group x time x measure *PSO.Ol. tPsO.OO1.
108 Public Health Nursing
Volume 9 Number 2 June 1992
nal consistency reliability at times 1, 2 , and 3. Both groups displayed increased knowledge from time 1 to time 2 and from time 1 to time 3. The same pattern of change was evident for the range of rest, relaxation, heat, and exercise activities, with both groups increasing the range of activities. Only the low depression group showed an increase in the total number of rest, relaxation, heat, and exercise activities reported from time 1 to time 2 . In summary, when depression was used t o separate the subjects who had participated in a SLE self-help course into two contrast groups, significant change was observed for the groups and for three interaction effects of group X time, group X measure, and group X time X measure.
DISCUSSION The findings of this study are consistent with theory of learned response to chronic illness. These participants did gain significantly in their level of enabling skill and self-efficacy, whether or not they were depressed on entering the classes. Those who were depressed at time of entry showed significant changes in five of seven learned response variables, perhaps indicating an evolving self-help response. Those who were not depressed maintained their entry-level of involvement in learned response, and did significantly increase the mediating variables of enabling skill and self-efficacy. Further work has to be done to evaluate which of the class activities contributes the most to initiating or maintaining a learned self-help response. The class-tied variables of SLE knowledge and the range and number of rest, relaxation, heat, and exercise activities did change significantly over time; however, other class-focused activities were not measured in this study. A current self-help-promoting intervention study with women receiving treatment for breast cancer that does measure specific intervention activities may offer insights about the effectiveness of other self-help class activities (Braden et al., 1990). The clinical implications of the findings concern the need for further development of community-based health programs for persons who are disabled or diagnosed with chronic illness. Results demonstrate that even those who have slowed or stopped their involvement in learning to live a healthier life can make significant gains in health-promoting behaviors. The success of this consumer-developed and -maintained program points the way for further development of a role for community health nurses as consultants to lay selfhelp groups organized to deal with chronic illness concerns.
REFERENCES Bakes, M. M. (1982). Environmental factors in dependency among nursing home residents: A social ecology analyses. In T. A. Wills, (Ed.), Basic processes in helping relationships (pp. 405-425). New York, Academic Press. Bandura, A. (1977). Self efficacy: Toward a unifying theory of behavior change. Psychological Review, 84, 191. Becker, H. A., Stuifbergen, A. K., Ingalski, K., & Sands, D. (1989). Health promoting attitudes and behaviors among persons with disabilities. International Journal Rehabilitation Research, 12(3), 235-250. Braden, C. J. (1990). A test of the self-help model: Learned response to chronic illness experience. Nursing Research, 39(1), 42-47. Braden, C. J. , Brodt-Weinberg,R., Depka, L., McClone, K . , & Tretter, S. (1987). Systemic lupus erythematosus (SLE) selfhelp course (leader’s manual). Atlanta, GA: Arthritis Foundation. Braden, C. J., Mishel, M. H., Longman, A., & Burns, L. (1990). A self-help intervention project (SHIP) in breast . . cancer treatment. Innovarions in Oncology Nursing, 6(2), 1, 8, 19. Bray, J. H., & Maxwell, S. E. (1989). Multivariate analysis of variance. Newburg Park, CA: Sage. Campbell, A., Converse, P., & Rogers, W. (1976). The quality of American life. New York: Sage. Chen, H. T. (1990). Theory-driven evaluations. Newburg Park, CA: Sage. DeJong, G., & Hughes, J. (1982). Independent living: Methodology for measuring long-term outcomes. Archives of’ Physical Medicine and Rehabilitation, 63, 68-73. Goodstadt, M. S., Simpson, R. I., & Loranger, P. L. (1987). Health promotion: A conceptual integration. American Journal of Health Promotion, 1(3), 58-63. Green, L. W. (1986). Individuals versus systems: An artificial distinction that divides and distorts. Health Link, 2(3), 29-30. Green, L., & Raeburn, J. (1990). Contemporary developments in health promotion: Definitions and challenges. In N. Bracht (Ed.), Health promotion at the community level (pp. 29-44). Newburg Park, CA: Sage. Haug, M., & Lavin, B. (1983). Consumerism in medicine: Challenging physician authority. Beverly Hills: Sage. Pender, N. J. (1987). Health promotion in nursing practice. Norwalk, CT: Appleton & Lange. Robinson, D. (1982). Self-help group in primary health care. World Health Forum, 2(2), 185-191. Rosenbaum, M. (1980). A schedule for assessing self-control behaviors: Preliminary findings. Behavior Therapy, 11, 109-121. Rosenbaum, M. (1988). Learned resourcefulness, stress and self-regulation. In S. Fisher, & J. Reason (Eds.), Handbook of life, stress, cognition and health (pp.. 483-496). New York: John Wiley & Sons. Seligman, M. E. P. (1975). Helplessness on depression, development and death. San Francisco: W. H. Freeman. Wright, B. A. (1960). Physical disability-A psychological upproach. New York: Harper & Row.
Description of Learned Response to Chronic Illness: Depressed versus Nondepressed Self-Help Class Participants Carrie Jo Braden, Ph.D., R.N.
Abstract The purpose of this study was to describe the effectiveness of a community-based health-promotion program that was collaboratively designed and that is currently being implemented by lay persons who have the diagnosis of systemic lupus erythematosus (SLE) and by health professionals. Two groups of SLE self-help course participants contributed data for this theory-guided preexperimental program impact study. Braden’s self-help model provided the theoretical framework. The variables addressed were severity of illness, limitations, uncertainty, enabling skill self-efficacy, self-worth, and life quality. The groups consisted of 35 subjects having low depression scores ( ~ 7 7 and ) 37 with high depression scores (a272) on a measure having a possible range of scores from 0 to 400. Data were collected at three points: at the beginning of class 1, after class 7 (7 wks later), and two months after completion of the course. A group (2) by time (3) by measures (10) MANOVA analysis procedure was used with a follow up univariate F test and Newman-Keuls multiple comparison procedure. Significant change was evident for the level of depression group and for three interaction effects of group X time, group X measure, and group X time X measure.
The purpose of this study was to describe the effectiveness of a community-based health-promotion program that was designed and implemented with persons who have the diagnosis of systemic lupus erythematosus (SLE). The description of program effectiveness addresses the learned response of two groups of participants in an SLE self-help course (Braden et al., 1987). One group represents the upper end of an adaptation continuum based on low depression scores and the other represents the lower end of the continuum based on high depression scores. Learned response variables included severity of illness, limitations, uncertainty, enabling skills, self-efficacy, self-worth, and life quality. Health promotion for persons with disabilities has received little notice, perhaps due to lack of focused attention on integrating polar views about system versus individual approaches, as well as lack of a framework for understanding the dynamics of how disabled persons engage in health-promoting strategies. Little research has been directed to health care attitudes and behaviors of persons with disabilities that could lead to interventions enhancing their health-promoting behaviors (Becker et al., 1989). Yet, if such behaviors are important for the nonimpaired population, they are more critical for those whose conditions leave them with a thinner margin of health (Becker et al., 1987; DeJong & Hughes, 1982). BACKGROUND
Several recent definitions provide a basis for viewing health promotion as an optimum vehicle for improving the health of persons with disabilities. Goodstadt, SimpResearch supported b y grant NCNR NRO 1686, from the Nason, and Loranger (1987), Pender (1987), and Green and tional Arthritis Foundation. Address correspondence to Carrie Jo Braden, University of Ar- Raeburn (1990) provided definitions from a perspective that places health professionals in the role of consultant, izona, College of Nursing, Tucson, A Z 85721.
103
104 Public Health Nursing
Volume 9 Number 2 June 1992
hypotheses from three alternative learning theories and incorporates the supported specifications from learned resourcefulness theory (Rosenbaum, 1988), learned helplessness theory (Seligman, 1975), and instrumental passivity theory (Baltes, 1982). The concepts addressed by this study of persons who participated in a health education program designed to enhance a learned self-help response to SLE are based on Braden’s self-help model construct specifications for 1) the stimulus for learning, 2) the negative reinforcers or adversities inherent in chronic illness, 3) the discriminant stimuli that act as “on” or “off” switches for involvement in learning, and 4) the outcome behaviors representing positive goals and serving as positive reinforcers for further learning. Robinson (1 982) defined a learned self-help response to negative life events as an informed process of facing definable, manageable adversities by maintaining control of everyday problems. The opposite response, succumbing, was described by Wright (1960), and addressed as learned helplessness by Seligman (1975) and as instrumental passivity by Baltes (1982). It can be defined as an uninformed process of withdrawing from definable, manageable adversities by giving in to everyday problems. Although it is known that most persons with a chronic illness do not give up and, in fact, learn by trial and error how to manage everyday problems themselves (Haug & Lavin, 1983), more remains to be known about how this learned self-help response works to turn the uncertainties and limitations of a disease like SLE into definable, manageable adversities. The self-help model poses that the perceived severity of the chronic illness initiates learning about living with it. The increased exposure to related adversities opens the door to a new learning frame. The model also hypothesizes that a repertory of enabling skills can serve as a cue to go ahead with discovering what to do on a day-to-day basis to reach desired goals. Some learning theorists, specifically Bandura (1977), assert that one’s THEORETICAL MODEL level of self-efficacy also mediates aversive aspects of A variety of theoretical models are also being used to illness. describe the dynamics of how persons engage in healthIt is also likely that negative mediators, for example, promoting behaviors despite chronic debilitating dis- depression, reduce involvement in learning. Depression ease. Braden’s self-help model is based on the assump- snuffs out a continued commitment to trying to overtions that learning is a fundamental process underlying come problems (Seligman, 1975). Maintenance of selfadaptation, that chronic illness represents exposure to worth and life quality are often goals targeted by pera new learning environment, and that learning poten- sons who are diagnosed with a chronic illness. Findings tially occurs within specific contexts involving the in- of Becker et al. (1989) indicated that more than half the teraction of intrapersonal, interpersonal, and environ- sample with disabilities had health-related goals that mental factors. The model has been used to describe focused on functioning well. However, not all perlearned self-help and learned helplessness response sons have a sufficient level of enabling skills or of selfamong persons with the diagnosis of rheumatoid arthri- efficacy to cue continued involvement in learning in tis or arthritis-related conditions (Braden, 1990). The order to reach desired goals, particularly in the face of self-help model emerged from the competitive testing of depression.
advocate, mediator, and supporter, and from an individual perspective that invests lay persons with decision power. These definitions, focusing on actualizing health potentials rather than on the eradicating evidence of disease or infirmity, concur that individuals diagnosed with chronic illness or living with a disability can be healthy. Emphasis is placed on self-care and mutual aid prerogatives, and provides for shared responsibility for health between individuals and systems (Green, 1986). A descriptive study of health-promoting behaviors among persons with disability demonstrated that many adults place great emphasis on the functional, adaptive, and self-actualizing aspects of health (Becker et al., 1989). Nearly three-fourths of the sample in this study perceived themselves as being healthy. Pender’s model of health promotion was applicable to persons with disabilities. A combination of attitudinal factors, such as self-efficacy, and demographic factors, such as age and sex, were related to self-reports of engaging in healthpromoting activities such as exercise, nutrition, and stress management (Becker et al., 1989). This study reports the results of a program that was based on the definition of health promotion as being the implementation of educational, organizational, economic, and environmental supports to maintain and enhance health (Goodstadt et al., 1987; Green & Raeburn, 1990; Pender, 1987). The program evaluation, a preexperimental design advocated by Chen (1990) as appropriate when the treatment (self-help course) directly acts on characteristics of the problem (a desire to achieve health), also incorporated lay persons in the evaluation planning process. Chen (1990) proposed theory-guided strategies for impact evaluation using the one group pretest-posttest design when costs of an experimental design are prohibitive. The preexperimental design provides worthwhile information about a program’s impact.
Braden: Enabling Skill, Self-care 105
DESCRIPTION OF THE INTERVENTION
tional Office in Atlanta, Georgia, and offered at 17 sites The self-help course incorporated the beliefs, needs, across the country. Class sites included four in Califorand concerns of persons with SLE by relying on nia and Texas, two in New Jersey and Kansas, and one consumer participation in developing the curriculum, each in Colorado, Oregon, Georgia, Iowa, and Illinois. rather than only on health professionals. The course Subjects were recruited into the study over four years uses lay teachers in partnership with health professional from early 1987 to late 1990. For attenders willing to participate in the study, data were collected at three teachers. points: during the first class session, after the last class The course activities focus on problem solving, cog(7 wks later), and two months after completing the nitive reframing, and enhancing belief in self. Exercise class. The Arthritis Foundation arranged for data coland relaxation activities were incorporated as methods of strengthening self-efficacy relative to specific health lection and mailing of completed forms to the invesbehavior. All activities are designed to enhance partic- tigator for data entry and analysis. Data on number of classes attended was available on 104 subjects. ipants’ interaction. The process of sharing with others The number ranged from one to seven classes attended is facilitated as a means of building self-worth as well as (mean 5.6 classes, SD 1.5). of adding to a repertory of problem-solving skills and to a set of alternative solutions for common problems. Demographics While trial-and-error over time may effectively strengthen enabling skills and self-efficacy and reduce depression Ninety-six percent (263) of the sample were women. in some persons, in others learning slows or even stops. Ages ranged from 21 to 83 years (mean 46 yrs). Most The goals of the self-help course were to restart trial- subjects were married (70%). Ethnic composition was and-error learning for participants who had slowed or 84% Caucasian, 7% black, 1.5% Asian, 0.7% American stopped involvement, and to maintain a self-help re- Indian, 2.9% Mexican, 0.4% Puerto Rican, 2.2% other, sponse in those who had continued to be involved in with 1% missing data. Fifty-nine percent had some collearning. Activities were designed to reduce a percep- lege education; only 7.4% had less than a high school tion of dependency, to facilitate uncertainty manage- diploma, with the lowest level of education being sevment, to strengthen enabling skills and self-efficacy, and enth grade. Forty-one percent worked outside the to build self-worth and subsequent improved life qual- home, with 30% of these indicating they were full-time ity. The theoretical basis for describing the learned employees. A nearly equal number (29%) indicated they response of participants specified the concepts of were full-time homemakers. Ten percent were retired perceived illness severity (stimulus), limitations and un- and 3% were seeking work. Fifteen percent reported certainty (negative reinforcers), enabling skill and self- they were disabled. The number of years ill ranged from efficacy (discriminant stimuli), and self-worth and life newly diagnosed to 30 years (mean 6.7 yrs). Twentyquality (desired outcomes). Class-tied variables in- four percent of the sample had been ill one year or cluded knowledge about SLE and activities to facilitate less. pain management, energy conservation, and mainteMeasures nance of functional status. Thus, the question addressed by this study was; what changes occur over time in per- The questionnaire was developed by health professionceived illness severity, limitations, uncertainty, enabling als with the help of persons having diagnosis of SLE skills, self-efficacy, self-worth, life quality, SLE knowl- who participated in establishing the original self-help edge, and the range and number of rest, relaxation, and course. The purpose was to put in place a procedure for exercise activities between participants who enter the collecting descriptive information about participants at class with high depression versus those who enter with the time the course was being distributed nationally. low depression? The developers met with several other persons, includMETHODS Procedure and Sample Human subjects criteria were met and approved by a human subjects review committee from a College of Nursing located in southwestern United States. A convenience sample of 273 subjects with diagnosis of SLE resulted from the recruitment of participants in a selfhelp course sponsored by the Arthritis Foundation Na-
ing this investigator, to collect instruments that had been used to measure various aspects of chronic illness. They then independently selected the items from the instruments they thought would be appropriate, and also created some of their own. The following variable definitions are consistent with variables and definitions provided in Braden’s (1990) self-help model. Severity of illness, one’s perceived level of infliction due to disease course characteristics, was measured by three visual analog (VAS) formatted items requesting in-
106 Public Health Nursing
Volume 9 Number 2 June 1992
formation about common SLE symptoms that had occurred during the past month. The 10-mm, horizontal VAS response line for the first item requesting response about muscle/joint pain had end point anchors of “as bad as pain could possibly be” and “no pain.” The end points for the second item response line were “no fatigue” and “as bad as fatigue could possibly be.” The third item regarding the amount of muscle/joint stiffness was anchored by “stiffness as bad as could be” and “no stiffness. High scores indicated greater symptom severity. The time 1 scores for the three summed items ranged from 0 to 281 (mean 157.5, SD 65, skewness -0.4), time 2 scores from 0 to 287 (mean 147.I , SD 60.2, - skewness 0.1), and times 3 from 0 to 285 (mean 143,1, SD 62, skewness 0.2). Internal consistency reliability estimates using Cronbach’s standardized alpha coefficient were 0.74, 0.81, and 0.79 for time 1, 2, and 3. Limitations, perceived level of inability to do things for oneself, was measured by a single item. The item asked respondents to select one of four statements that best describes what they are able to do. The four statements were “I can do everything I want to do”; “I can do most of the things I want to do, but have some limitations”; “I can do some, but not all, of the things I want to do”; and “I have many limitations and I can do hardly any of the things I want to do.” The item was scored with the highest score, 4, indicating greatest limitation. At time 1 the mean was 2.3 (SD 0.7, skewness 0.2), at time 2 it was 2.2 (SD 0.6, skewness 0.5), and at time 3 it was 2.2 (SD 0.6, skewness 0.2). Uncertainty, one’s perceived amount of ambiguity about treatment effectiveness, was measured by three items, each having a VAS response line with end points of “not at all certain” and “very certain.” The items concerned certainty in ability to use self-care techniques, effectiveness of medications to improve SLE, and ability to decrease pain or stiffness. High scores indicated low levels of certainty. Summed scores at time 1 ranged from 0 to 282 (mean 128.6, SD 66.5, skewness 0.2), at time 2 from 1 to 300 (mean 96.9, SD 62.7, skewness 1 . 1 ) , and at time 3 from 0 to 300 (mean 98.7, SD 66.7, skewness 0.8). The standardized Cronbach’s alpha for times 1 , 2, and 3 were 0.66, 0.79, and 0.79. Depression, one’s level of despondency, was measured by four items having a VAS response format. The response line for the first item, “how have you been feeling in general?”, was anchored with “in excellent spirits” and “in very low spirits.” The response line for the second item, “have you felt so sad, discouraged, hopeless, or had so many problems that you wondered if anything was worthwhile?” was anchored by “extremely so, to the point I have just about given up” and “not at all.” The response line for the third item, “have you felt down hearted and blue?”, was anchored by “all ”
of the time” and “none of the time.” The last item asking, “how depressed have you been?” had response line end points of “very depressed” and “not at all depressed.” The items were scored in the direction of depression. Time 1 depression scores ranged from 0 to 381 (mean 174.6, SD 97.8, skewness 0.2), time 2 scores from 0 to 400 (mean 147.5, SD 92.7, skewness 0.6), and time 3 from 0 to 396 (mean 153.2, SD 97.5, skewness 0.4). Internal consistency estimated by the standardized Cronbach’s alpha was 0.90 for time 1 , 0.86 for time 2, and 0.92 for time 3. Enabling skill, one’s perceived level of ability to manage adversity, was measured by six items drawn from Rosenbaum’s (1980) 36-item self-control schedule. A VAS response format with end points of “true about me” and “not true about me” was used. Items addressed cognitive reframing skills: “often by changing my way of thinking, I am able to change my feelings about almost anything”; “when I am depressed I try to keep myself busy with things 1 like to do”; and “when I feel pain, I try to divert my thoughts from it.” Items also addressed problem-solving skills: “when faced with a difficult problem, I try to approach its solution in a systematic way”; and “I keep track of how well a treatment works for me and I usually plan my work when I’m faced with a lot of things to do.” The items were scored in a positive direction for high-enabling skill. At time 1 the summed score on the six items ranged from 0 to 600 (mean 393.3, SD 129.1, skewness -0.9), at time 2 from 33 to 600 (mean 416.3, SD 117, skewness - 1 . I ) , and at time 3 from 0 to 600 (mean 43 -, SD 116.7, skewness - 1.2). The standardized Cronbach’s alpha at time 1 was 0.82, at time 2 0.85, and at time 3 0.87. Self-efficacy, the strength of conviction that one can do what is necessary to control a primary SLE-related symptom, was measured by a single item having a VAS response format. The response line for the item, “how satisfied are you with your ability to control fatigue?” was anchored by “very dissatisfied” and “very satisfied.” Items were scored in a positive direction for greater self-efficacy. Scores on the item ranged from 0 to 100 (mean 39.6, SD 28.3, skewness 0.5) for time 1 , from 0 to 100 (mean 52.8, SD 24.8, skewness-0.2), at time 2, and from 0 to 100 (mean 56.3, SD 25.3, skewness -0.4) at time 3. Self-worth, level of positive feelings about one’s own being, was measured by 10 items having a VAS response format with end points of “true about me” and “not true about me.” Items concerned ways subjects felt about themselves in general. They were scored in a positive direction for self-worth. Scores at time 1 ranged from 0 to 1000 (mean 629.9, SD 223.6, skewness-0.4), at time 2 from 5 to 1000 (mean 682.8, SD 217.4, skewness - l . l ) , and at time 3 from 0 to 1000 (mean 667.4,
Braden: Enabling Skill, Self-care 107
SD 232.7, skewness -0.8). The standardized Cronbach’s alpha at time 1 was 0.90; at time 2 was 0.92, and at time 3 was 0.93. Life quality, perceived level of satisfaction with one’s personal situation, was measured by three items drawn from Campbell, Converse, and Rogers’s (1976) 10-item semantic differential scale, inventory of well-being. The items selected were “my present life is enjoyable-miserable,” or “full-empty,” and “in thinking about my life as a whole, I am completely satisfied-completely dissatisfied.” Higher scores reflected greater life quality. Responses on the three summed items at time 1 ranged from 0 to 300 (mean 173.4, SD 67.1, skewness -0.3), at time 2 from 0 to 300 (mean 187.4, SD 63.8, skewness - 0.4), and at time 3 from 2 to 300 (mean 184.6, SD 68.6, skewness - 0.4). The standardized Cronbach’s alpha coefficient for time 1 was 0.85, for time 2 was 0.85, and for time 3 was 0.90. Three additional variables, SLE knowledge, activity number, and activity range, were included to address information gained through class content and through rest, relaxation, heat, and exercise alternatives that were incorporated into class activities for pain management and for strengthening self-efficacy. Knowledge about SLE, the level of information about disease characteristics and treatment alternatives, was measured by seven multiple-choice items. The range of correct answers at time 1 was 0 to 7 (mean 4.7, SD 1.5, skewness -0.6), at time 2 from 1 to 7 (mean 5.5, SD 1.2, skewness -0.7), and at time 3 from 1 to 7 (mean 5.5, SD 1.2, skewness - 0.9). Internal consistency reliability as estimated by Cronbach’s alpha was 0.47 at time 1 , 0.45 at time 2, and 0.54 at time 3. Rest, relaxation, heat exercise range, the number of different types of pain-management and self-strengthening activities engaged in during one week, was measured by counting the number of different types of activities reported from a check list. Subjects also listed additional physical activities engaged in during a week’s time. These were categorized, and the number of different activity categories added to the reported number of activities from the provided list. The range of activities engaged in at time 1 was 0 to 10 (mean 3.3, SD 1.8, skewness O S ) , at time 2 0 to 10 (mean 3.9, SD 1.9, skewness - 0. l), and at time 3 0 to 1 1 (mean 4.2, SD 1.9, skewness 0.8). Rest, relaxation, heat, exercise number, the total number of times pain-management and self-strengthening activities were engaged in during a week, was measured by summing the reported number of times for each activity. At time 1 the number of times activities were engaged in ranged from 0 to 91 (mean 13.6, SD 11.2, skewness 2) and at time 3 from 0 to 70 (mean 16.3, SD 10.8, skewness 1.5).
RESULTS A group (2) time (3) by measures (10) MANOVA procedures was used to answer the question about the pattern of change in learned response variables and in class-tied variables occurring over time in two groups, those having high depression scores at time of class entry and those having low depression scores. The MANOVA procedure was selected because of the desire to investigate a variable, depression, that could contribute most to separation of learned response patterns (Bray & Maxwell, 1989, p. 11). The depression groups were formed from subjects having depression scores above or equal to a + 1 SD (37 subjects with mean scores 2272) and those with scores below or equal to a - 1 SD (35 subjects with mean scores 677). The results are presented in Table 1. Significant changes were found between high depression and low depression as a main effect. Three significant interaction effects were found for depression group and time, measure, and time and measure. The univariate F test and Newman-Keuls multiple comparison procedure using P =s0.05 were selected to describe patterns of change within each group. These procedures revealed seven sources of significant changes in the high depressed group and five in the low depressed group. No significant changes were found for either group over time for severity of illness or limitation. Uncertainty changed significantly only for the high depression group, with a reduction in uncertainty occurring between times 1 and 2 and times 1 and 3. Both groups changed significantly over time, with enabling skill, self-efficacy, and self-worth increasing between times 1 and 2 and times 1 and 3 for the high depression group. The low depression group evidenced significantly more enabling skill, self-efficacy, and self-worth only from time 1 to time 3. Life quality increased significantly over time only for those in the high depression group. The comparisons for SLE knowledge have to be viewed with caution, given the low estimates for interTABLE 1. Multivariate Analysis of Variance of Depression Group by Time and Measure
Source of Variation
df
F
Group Time
1170 21140 91630 21140 91630 1811264 1811 264
8.3* 0.9 0.3 30.lt 20.2t 0.3 8.9t
Measure Group x time Group x measure Time x measure Group x time x measure *PSO.Ol. tPsO.OO1.
108 Public Health Nursing
Volume 9 Number 2 June 1992
nal consistency reliability at times 1, 2 , and 3. Both groups displayed increased knowledge from time 1 to time 2 and from time 1 to time 3. The same pattern of change was evident for the range of rest, relaxation, heat, and exercise activities, with both groups increasing the range of activities. Only the low depression group showed an increase in the total number of rest, relaxation, heat, and exercise activities reported from time 1 to time 2 . In summary, when depression was used t o separate the subjects who had participated in a SLE self-help course into two contrast groups, significant change was observed for the groups and for three interaction effects of group X time, group X measure, and group X time X measure.
DISCUSSION The findings of this study are consistent with theory of learned response to chronic illness. These participants did gain significantly in their level of enabling skill and self-efficacy, whether or not they were depressed on entering the classes. Those who were depressed at time of entry showed significant changes in five of seven learned response variables, perhaps indicating an evolving self-help response. Those who were not depressed maintained their entry-level of involvement in learned response, and did significantly increase the mediating variables of enabling skill and self-efficacy. Further work has to be done to evaluate which of the class activities contributes the most to initiating or maintaining a learned self-help response. The class-tied variables of SLE knowledge and the range and number of rest, relaxation, heat, and exercise activities did change significantly over time; however, other class-focused activities were not measured in this study. A current self-help-promoting intervention study with women receiving treatment for breast cancer that does measure specific intervention activities may offer insights about the effectiveness of other self-help class activities (Braden et al., 1990). The clinical implications of the findings concern the need for further development of community-based health programs for persons who are disabled or diagnosed with chronic illness. Results demonstrate that even those who have slowed or stopped their involvement in learning to live a healthier life can make significant gains in health-promoting behaviors. The success of this consumer-developed and -maintained program points the way for further development of a role for community health nurses as consultants to lay selfhelp groups organized to deal with chronic illness concerns.
REFERENCES Bakes, M. M. (1982). Environmental factors in dependency among nursing home residents: A social ecology analyses. In T. A. Wills, (Ed.), Basic processes in helping relationships (pp. 405-425). New York, Academic Press. Bandura, A. (1977). Self efficacy: Toward a unifying theory of behavior change. Psychological Review, 84, 191. Becker, H. A., Stuifbergen, A. K., Ingalski, K., & Sands, D. (1989). Health promoting attitudes and behaviors among persons with disabilities. International Journal Rehabilitation Research, 12(3), 235-250. Braden, C. J. (1990). A test of the self-help model: Learned response to chronic illness experience. Nursing Research, 39(1), 42-47. Braden, C. J. , Brodt-Weinberg,R., Depka, L., McClone, K . , & Tretter, S. (1987). Systemic lupus erythematosus (SLE) selfhelp course (leader’s manual). Atlanta, GA: Arthritis Foundation. Braden, C. J., Mishel, M. H., Longman, A., & Burns, L. (1990). A self-help intervention project (SHIP) in breast . . cancer treatment. Innovarions in Oncology Nursing, 6(2), 1, 8, 19. Bray, J. H., & Maxwell, S. E. (1989). Multivariate analysis of variance. Newburg Park, CA: Sage. Campbell, A., Converse, P., & Rogers, W. (1976). The quality of American life. New York: Sage. Chen, H. T. (1990). Theory-driven evaluations. Newburg Park, CA: Sage. DeJong, G., & Hughes, J. (1982). Independent living: Methodology for measuring long-term outcomes. Archives of’ Physical Medicine and Rehabilitation, 63, 68-73. Goodstadt, M. S., Simpson, R. I., & Loranger, P. L. (1987). Health promotion: A conceptual integration. American Journal of Health Promotion, 1(3), 58-63. Green, L. W. (1986). Individuals versus systems: An artificial distinction that divides and distorts. Health Link, 2(3), 29-30. Green, L., & Raeburn, J. (1990). Contemporary developments in health promotion: Definitions and challenges. In N. Bracht (Ed.), Health promotion at the community level (pp. 29-44). Newburg Park, CA: Sage. Haug, M., & Lavin, B. (1983). Consumerism in medicine: Challenging physician authority. Beverly Hills: Sage. Pender, N. J. (1987). Health promotion in nursing practice. Norwalk, CT: Appleton & Lange. Robinson, D. (1982). Self-help group in primary health care. World Health Forum, 2(2), 185-191. Rosenbaum, M. (1980). A schedule for assessing self-control behaviors: Preliminary findings. Behavior Therapy, 11, 109-121. Rosenbaum, M. (1988). Learned resourcefulness, stress and self-regulation. In S. Fisher, & J. Reason (Eds.), Handbook of life, stress, cognition and health (pp.. 483-496). New York: John Wiley & Sons. Seligman, M. E. P. (1975). Helplessness on depression, development and death. San Francisco: W. H. Freeman. Wright, B. A. (1960). Physical disability-A psychological upproach. New York: Harper & Row.
