HEALTH PSYCHOLOGY, 1992,11(4), 257-261 Copyright © 1992, Lawrence Erlbaum Associates, Inc.
Assessing Motivational Readiness and Decision Making for Exercise Bess H. Marcus The Miriam Hospital and Brown University School of Medicine
William Rakowski Department of Community Health Center for Gerontology and Health Care Research Brown University
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Joseph S. Rossi Cancer Prevention Research Center University of Rhode Island Motivational and cognitive processes of behavior change with respect to the area of exercise adoption were investigated. A total of 778 men and women, recruited from four worksites, answered a 40-item questionnaire consisting of statements based on constructs from the transtheoretical model of behavior change. Principal-components analysis identified two factors—one a 6-item component representing avoidance of exercise (Cons), the other a 10-item component representing positive perceptions of exercise (Pros). Analysis of variance showed that the Pros, Cons, and a Decisional Balance measure (Pros minus Cons) were significantly associated with stage of exercise adoption. Results are consistent with applications of the model to smoking cessation and other areas of behavior change. Distinctions between exercise adoption and behaviors such as smoking cessation, weight loss, and alcoholism are discussed. Key words: exercise, decisional balance, health behavior, transtheoretical model, behavioral adoption
The benefits of exercise influence a wide variety of diseases and conditions affecting both physical and mental health. Regular exercise can help prevent and treat coronary heart disease, osteoporosis, diabetes, hypertension, and depression (Harris & Associates, 1989). Because the benefits of exercise are clear, researchers and clinicians are faced with two main challenges: first, how to get people to initiate exercise behavior and, second, how to help active people maintain their exercise behavior. Little is known about how to intervene in a population so as to increase and then sustain the proportion of individuals interested in adopting regular exercise. More is known about relapse after a program has been started. Research has documented that, across a variety of populations, approximately 50% of individuals who join an exercise program will drop out during the first 3 to 6 months (Carmody, Senner, Manilow, & Matarazzo, 1980; Dishman, 1988b). Although many studies have been conducted to address the problem of exercise relapse, and some have shown positive effects on shortterm adherence to relatively brief exercise programs, little success has been attained in improving long-term maintenance of exercise behavior (Dishman, 1982,1988a, 1988b; Martin & Dubbert, 1982, 1984). This pattern of exercise relapse is similar to the negatively accelerated relapse curve often seen in the addictions (Hunt, Barnett, & Branch, 1971). This study examined the potential for extending a theory of behavior change based on cognitive and social learning constructs to the area of exercise adoption. The transtheoretical model was proposed initially by Prochaska and DiClemente (1982,1983) as a general model of intentional behavior change. The core of the model is a sequence of stages along a continuum of behavioral change from
Requests for reprints should be sent to Bess H. Marcus, Division of Behavioral Medicine, RISE Building, The Miriam Hospital and Brown University School of Medicine, 164 Summit Avenue, Providence, RI 02906.
precontemplation (not intending to make changes), to contemplation (considering a change), to preparation (making small changes), to action (actively engaging in the new behavior), to maintenance (sustaining the change over time). There is also the possibility of relapsing back to an earlier stage, as change is hypothesized to occur in a cyclical rather than linear pattern. The model also defines a set of outcome or intermediate variables that includes decisional balance—the pros and cons of behavior change (Velicer, DiClemente, Prochaska, & Brandenburg, 1985). Decisional balance is based on the theoretical model of decision making developed by Janis and Mann (1977) and was initially applied to smoking cessation by Velicer et al. (1985). It is based on a comparison of the perceived positive aspects (pros) and negative aspects (cons) of a new behavior. DiClemente et al. (1991) described the difference on each of these measures across the stages of change for smoking cessation. The stages of smoking behavior have been found to correspond to differences in the balance of the perceived pros and cons of that behavior on the decisional balance measure. Persons in action and maintenance have a decisional balance favoring the positive features (pros), persons in precontemplation have a balance reflecting reasons not to change (cons), and persons in contemplation tend to fall between those in precontemplation and action. In longitudinal studies of smoking cessation, decisional balance has been especially useful in predicting movement from the precontemplation to the contemplation stage and in predicting behavior change (Prochaska, Velicer, DiClemente, Guadagnoli, & Rossi, 1991). The transtheoretical model has been applied most thoroughly to smoking cessation (DiClemente et al., 1991; Prochaska & DiClemente, 1983; Prochaska, Velicer, DiClemente, & Fava, 1988). However, it has more recently been extended to other health-related behaviors including exercise (Marcus, Banspach et al., in press; Marcus & Owen, 1992; Marcus, Rossi, Selby, Niaura, & Abrams, in press; Marcus, Selby, Niaura, & Rossi, 1992; Marcus & Simkin,
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
258
MARCUS, RAKOWSKI, ROSSI
in press), alcohol use (DiClemente & Hughes, 1990), weight loss (O'Connell & Velicer, 1988), screening mammography (Rakowski et al., 1992), and ultraviolet light exposure (Rossi, 1989a, 1989b) and may also be helpful for understanding stability and change at the level of organizations (Abrams, Emmons, Niaura, Goldstein, & Sherman, 1991). A strength of this model is its focus on the dynamic nature of health behavior change. Because this is a dynamic model, the different transitions in adoption and maintenance of a behavior described by other researchers (Dishman, 1982; Sallis & Hovell, 1990; Sonstroem, 1988) can be specifically examined. This model suggests that behavior change is not an all-or-none phenomenon and that individuals who stop performing a behavior may have intentions to start again (Sonstroem, 1988). The present article reviews questionnaire development pertinent to the decisional balance aspect of the transtheoretical model for extension of the model to exercise adoption. Our particular interests were in whether scales representing pros and cons could be developed (measurement development) and whether these scales would be associated with stages of exercise adoption (model testing) as a basis for further model development in this area of health promotion and disease prevention. METHOD
Subjects Subjects were recruited as part of a worksite health promotion project emphasizing health-risk appraisal and smoking cessation. Seven hundred seventy-eight of a possible 1,173 (66%) male and female employees of four worksites—a retail outlet, an industrial manufacturer, a government agency, and a medical center—participated. Fifty-four percent of the sample was female, average age was 41.5 years (SD = 11.0 years), and average years of education was 13.5 (SD = 2.0 years). Ninety-five percent of the subjects were White, 70% were married, and 17% were currently smoking. Median annual household income was $35,000, with 70% of employees involved in white-collar occupations. Procedure The present data were collected in the course of normal follow-up surveying at the worksites. Employees received a cover letter from their chief executive officer explaining that they were invited to continue their participation in a study about health promotion at the worksite. Persons who volunteered for the study completed questionnaires on exercise and basic demographic information and had previously provided informed consent. Respondents also completed questionnaires about their smoking status and about other lifestyle behaviors as part of a large study on health-risk status and smoking behavior at the workplace. Details concerning the larger study will be reported elsewhere. Subjects were informed that their names would be entered into a drawing for a $100 prize in return for their participation. Measures Decisional Balance. Development of the Decisional Balance measure for exercise followed the sequential method of scale construction described by Jackson (1970,1971; see also Comrey, 1988)
and involving a sequence of steps to ensure content and internal validity. An initial pool of approximately 75 statements reflecting the positive (pros) and negative (cons) aspects of exercise was generated by a small group of male and female regular exercisers and non-exercisers, including researchers and laypersons. The statements were reviewed and revised to improve clarity of expression, eliminate redundancies, and ensure representation among the decisional categories recommended by Janis and Mann (1977): (a) gains or losses expected for oneself, (b) gains or losses expected for significant others, (c) self-approval or -disapproval due to the behavior, and (d) approval or disapproval by others due to the behavior. An example of a pro item was "I would feel healthier if I exercised regularly", and an example of a con item was "I do not like the way my body feels when I exercise." The final pool of items consisted of 40 statements, including 20 pros and 20 cons of exercising. Subjects were asked to indicate, on a 5-point Likert scale ranging from not at all important (1) to extremely important (5), how important each statement was with respect to their decision to exercise or not to exercise.
Stages of adoption for exercise. Stages of adopting exercise were measured using an 11-point scale in the shape of a ladder. Each rung had a number (0 through 10), and 5 rungs also had written labels to serve as anchor points. The 0 rung was labeled / currently do not exercise and I do not intend to start exercising in the next 6 months (precontemplation), 2 was labeled / currently do not exercise, but I am thinking about starting to exercise in the next 6 months (contemplation), 5 was labeled / currently exercise some, but not regularly (preparation), 8 was labeled I currently exercise regularly but I have only begun doing so within the last 6 months (action), and 10 was labeled / currently exercise regularly and have done so for longer than 6 months (maintenance). The formal stage-of-adoption labels (just noted in parentheses) were not placed on the ladder. Respondents picked the rung on the ladder that most accurately described their current exercise behavior. Regular exercise was defined as exercising three or more times per week for at least 20 min each time (American College of Sports Medicine, 1990). Respondents could indicate numbers other than those that were labeled, but the labels corresponded to the minimum requirements for membership in a particular stage. Thus, a subject responding with a "4" on the ladder was classified as a contemplator because the minimum requirements for the preparation stage had not been met. In practice, almost 80% of respondents selected one of the five labeled response categories, suggesting that the stage ladder could be shortened in future studies. Reliability of the stages-of-exercise-adoption measure has been examined. The Kappa index of reliability over a 2-week period was .78 (N = 20; Marcus et al., 1992).1 Concurrent validity for this measure has been demonstrated by its significant association with the Seven Day Recall Physical Activity Questionnaire (Marcus & Simkin, in press). RESULTS Principal-Components Analysis Seventeen items (3 Pros, 14 Cons) from the questionnaire were eliminated due to excessive positive or negative response, which 1
Fleiss (1981) reported that values of Kappa above .75 indicate strong agreement.
259
ADOPTION OF EXERCISE
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
limited the variance of the items. The remaining 23 items were subjected to a principal-components analysis with a varimax rotation. The number of components to be retained was determined by the scree method (Cattell, 1966) and interpretability. After rotation, an item was judged to be salient for a component if there was a component loading of .50 or greater and if the item did not load on another component. TABLE 1 Results of Principal-Components Analysis for Statements Representing Positive and Negative Orientations Toward Exercise (Pros and Cons) Component Loading
Statement Wording Pros I would have more energy for my family and friends if I exercised regularly. Regular exercise would help me relieve tension. I would feel more confident if I exercised regularly. I would sleep more soundly if I exercised regularly. I would feel good about myself if I kept my commitment to exercise regularly. I would like my body better if I exercised regularly. It would be easier for me to perform routine physical tasks if I exercised regularly. I would feel less stressed if I exercised regularly. I would feel more comfortable with my body if I exercised regularly. Regular exercise would help me have a more positive outlook on life. Cons I think I would be too tired to do my daily work after exercising. I would find it difficult to find an exercise activity that I enjoy that is not affected by bad weather. I feel uncomfortable when I exercise because I get out of breath and my heart beats very fast. Regular exercise would take too much of my time. I would have less time for my family and friends if I exercised regularly. At the end of the day, I am too exhausted to exercise.
.77 .81 .86
Two components were interpreted from the varimax pattern. The first component was composed of 6 items that highlighted unfavorable perceptions of exercise and reasons not to exercise. Conceptually, this factor represented the Cons dimension evident in previous work with the transtheoretical model. The second factor was composed of 10 items. This second category represented a positive or pro perception of exercise. The 6 Cons and 10 Pros items were subjected to another principal-components analysis in order to determine their final loadings on their respective components. The two components accounted for 60.4% of the total item variance. The loadings for each of the 16 items are provided in Table 1. Each scale was replicated in the second analysis (i.e., all loadings were greater than .50, and no other components emerged). Coefficient alpha (internal consistency) reliabilities were satisfactory (Cons = .79, Pros = .95). Stages of Adoption
.72 .86 .81 .84 .83 .85 .85
.71 .52
.62
.79 .70 .78
One-way analysis of variance was used to examine the association among stages of exercise adoption and the Pros and Cons indices. In order to provide a standard metric, the Pros and Cons indices were converted to T scores (M = 50, SD = 10). In addition, the Decisional Balance measure was created by subtracting Cons from Pros. Table 2 presents the T-score means and standard deviations by stage of exercise adoption. Differences on the Pros scale by stage of adoption were significant, F(4, 712) = 34.84, p < .0001, t]2 = .16. Based on a NewmanKeuls analysis, Pros scores were significantly higher for subjects in maintenance compared to precontemplation and contemplation; action compared to precontemplation, contemplation, and preparation; and preparation and contemplation compared to precontemplation. Therefore, 7 of the 10 possible pairwise contrasts were significant. Differences on the Cons scale by stage of adoption were also significant, F(4, 710) = 25.31, p < .0001, T)2 = .12. Based on a Newman-Keuls analysis, Cons scores were significantly lower for subjects in maintenance compared to action; action and maintenance compared to preparation; maintenance compared to precontemplation; and all stages compared to contemplation. Thus, 8 of the 10 possible pairwise contrasts were significant. Last, the Decisional Balance analysis also achieved significance, F(4, 709) = 49.32, p < .0001, v\2 = .22. Newman-Keuls follow-up indicated that all stages were significantly different from all other stages, with precontemplators scoring lowest and maintainers scoring highest. Figure 1 presents a pictorial view of the Pros and Cons scales by stage of exercise adoption.
TABLE 2 Means and Standard Deviations of the Pros, Cons, and Decisional Balance Scales by Stage of Exercise Adoption Stage of Adoption Precontemplation^ Scale
M
Pros Cons Decisional Balance
36.75 50.46
a
SD
c
11.13 d
e
Preparation0
Action6
Maintenance"
SD
M
SD
M
SD
M
SD
49.56 53.79
8.95 9.25
50.72 50.32
8.78 9.01
52.97 47.97
7.66 9.44
53.02 44.10
9.89 8.49
-4.21
11.16
0.40
11.97
5.00
12.21
9.09
13.12
M
11.33 11.93
-13.73 b
Contemplation^
n = 53. n = 242. n = 182. n = 101. n = 139.
260
MARCUS, RAKOWSKI, ROSSI
60
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
50
40
30
PreContemplation
Cuntemplation
Preparation
Action
Maintenance
FIGURE 1 T-score means for the Pros and Cons scales by stage of exercise adoption.
DISCUSSION This investigation provided encouraging results for extending the transtheoretical model of behavior change to the area of exercise adoption. Three important objectives were supported. First, a pool of items was developed guided by the model and yielded principal components consistent with major decisional balance constructs. This outcome was important for measurement development. Second, the decisional balance measures reflected hypothesized differences across stages of exercise adoption. This outcome was important for model testing. Last, and due to these two points, there is reason to pursue additional model development with exercise. The generalizability of these findings is limited in that they are based on a single cross-sectional investigation utilizing self-report data with no objective information on actual exercise behavior. However, the present results do provide some evidence that participants' reports of current exercise behavior (i.e., stages of adoption) correspond to beliefs about favorable and unfavorable features (i.e., pros, cons, decisional balance). Conversely, knowing participants' beliefs may portend the degree of acceptance or reluctance encountered by attempts to produce behavioral change toward regular exercise behavior. That the transtheoretical model appears to translate to exercise is encouraging, because exercise adoption involves the initiation of a positive behavior for health promotion and disease prevention in a general population rather than the elimination of a potentially harmful behavior like smoking or the reduction of a clinically assessed condition (e.g., obesity).
The 16 items eventually used from the present instrument provided indices with adequate empirical properties and conceptual fit to the model. At the same time, the items were only a subset of our pool, and there is certainly ample room for other investigators to test instruments of their own design. Replication of the pattern of statistical significance found here, but using other instruments, could add further support for the extension of the model to exercise. Additional investigations examining the reliability and validity of these measures are warranted. Because intervention is the ultimate objective of the model, an important consideration is that there be strong evidence in favor of the efficacy of the newly targeted health behavior. The consensus for the benefits of smoking cessation and exercise adoption make them appropriate foci for the model. If the present cross-sectional findings hold up in longitudinal studies of exercise behavior, it will be essential for interventions to be aimed at increasing participants' pros of exercise in order for movement through the stages to be accomplished. Although helping participants to decrease their cons of exercise is also important, this alone is not likely to result in enhanced readiness to be active. Knowledge of participants' pro and con beliefs in addition to the sociodemographic, biologic, and environmental influences on them (Dishman, 1982; Martin & Dubbert, 1982) will enhance the ability of researchers and clinicians to design effective exercise promotion interventions. An exploration of specific Pros and Cons items reveals many themes in common with previous empirical observations regarding the determinants of exercise. Pros items focusing on enhanced
ADOPTION OF EXERCISE confidence, feeling good about oneself, and having more energy for one's family and friends support the determinants-related work of Dishman, Sallis, and Orenstein (1985) and Sallis and Hovell (1990). Cons items such as being too tired to exercise, being concerned about bad weather, feeling uncomfortable and out of breath, and not having enough time also support previous research on the determinants of exercise. The transtheoretical model provides a framework for integrating these diverse variables and creating an empirical index that can be used to better understand the determinants of exercise and therefore to design more efficacious exercise interventions.
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ACKNOWLEDGMENTS This research was partially supported by National Institutes of Health Grant SO7RR05818 and National Cancer Institute Grant CA50087. We thank Elaine Taylor, Sheilarae Carpentier, Anna Klevak, Nancy Keenan, and Mark Morgenstern for their efforts on this project.
REFERENCES Abrams, D. B., Emmons, K. M., Niaura, R. S., Goldstein, M. G., & Sherman, C. B. (1991). Tobacco dependence. In P. E. Nathan, J. W. Langenbucher, B. S. McCrady, & W. Frankenstein (Eds.), The annual review of addictions treatment and research (Vol. 1, pp. 391—436). New York: Pergamon. American College of Sports Medicine. (1990). Position statement on the recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness in healthy adults. Medicine and Science in Sports and Exercise, 22, 265-274. Carmody, T. P., Senner, J. W., Manilow, M. R., & Matarazzo, J. D. (1980). Physical exercise rehabilitation: Long-term dropout rate in cardiac patients. Journal of Behavioral Medicine, 3, 163—168. Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276. Comrey, A. L. (1988). Factor-analytic methods of scale development in personality and clinical psychology. Journal of Consulting and Clinical Psychology, 56, 754-761. DiClemente, C. C, & Hughes, S. (1990). Stages of change profiles in outpatient alcoholism treatment. Journal ofSubstance Abuse, 2,217-235. DiClemente, C. C , Prochaska, J. O., Fairhurst, S., Velicer, W. F., Velasquez, M., & Rossi, J. (1991). The process of smoking cessation: An analysis of precontemplation, contemplation, and preparation stages of change. Journal of Consulting and Clinical Psychology, 59, 295-304. Dishman, R. K. (1982). Compliance/adherence in health-related exercise. Health Psychology, 1, 237-267. Dishman, R. K. (1988a). Exercise adherence research: Future directions. American Joural of Health Promotion, 3, 52—56. Dishman, R. K. (1988b). Overview. In R. Dishman (Ed.), Exercise adherence (pp. 1-9). City, IL: Human Kinetics. Dishman, R. K., Sallis, J. F., & Orenstein, D. R. (1985). The determinants of physical activity and exercise. Public Health Reports, 100, 158-172. Fleiss, J. L. (1981). Statistical methods for rates and proportions (2nd ed.). New York: Wiley. Harris, L, & Associates, Inc. (1989). The prevention index '89: summary report. Emmaus, PA: Rodale. Hunt, W. A., Barnett, L. W., & Branch, L. G. (1971). Relapse rates in addictions programs. Journal of Clinical Psychology, 27, 455-456.
261
Jackson, D. N. (1970). A sequential system for personality scale development. In C. D. Spielberger (Ed.), Current topics in clinical and community psychology (Vol. 2, pp. 61-96). New York: Academic. Jackson, D. N. (1971). The dynamics of structured personality tests. Psychological Review, 78, 229-248. Janis, I. L., & Mann, L. (1977). Decision making: A psychological analysis of conflict, choice and commitment. New York: Free Press. Marcus, B. H., Banspach, S. W., Lefebvre, R. L., Rossi, J. S., Carleton, R. A., & Abrams, D. B. (in press). Using the stages of change model to increase the adoption of physical activity among community participants. American Journal of Health Promotion. Marcus, B. H., & Owen, N. (1992). Motivational readiness, self-efficacy and decision-making for exercise. Journal ofApplied Social Psychology, 22, 3-16. Marcus, B. H., Rossi, J. S., Selby, V. C , Niaura, R. S., & Abrams, D. B. (in press). The stages and processes of exercise adoption and maintenance in a worksite sample. Health Psychology. Marcus, B. H., Selby, V. C , Niaura, R. S., & Rossi, J. S. (1992). Self-efficacy and the stages of exercise behavior change. Research Quarterly for Exercise and Sport, 63, 60—66. Marcus, B. H., & Simkin, L. R. (in press). The stages of exercise behavior. Journal of Sports Medicine and Physical Fitness. Martin, J. E., & Dubbert, P. M. (1982). Exercise applications and promotion in behavioral medicine: Current status and future directions. Journal of Consulting and Clinical Psychology, 50, 1004-1017. Martin, J. E., & Dubbert, P. M. (1984). Behavioral management strategies for improving health and fitness. Journal of Cardiac Rehabilitation, 4, 200-208. O'Connell, D., & Velicer, W. F. (1988). A decisional balance measure for weight loss. International Journal of Addictions, 23, 729—750. Prochaska, J. O., & DiClemente, C. C. (1982). Transtheoretical therapy: Toward a more integrative model of change. Psychotherapy: Theory, Research, andPractice, 20, 161-173. Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self change of smoking: Toward an integrative model. Journal of Consulting and Clinical Psychology, 51, 390-395. Prochaska, J. O., & DiClemente, C. C. (1985). Common processes of self-change in smoking, weight control, and psychological distress. In S. Shiftman & T. Wills (Eds.), Coping and substance use (pp. 345-363). New York: Academic. Prochaska, J. O., Velicer, W. F., DiClemente, C. C , & Fava, J. (1988). Measuring processes of change: Applications to the cessation of smoking. Journal of Consulting and Clinical Psychology, 56, 520—528. Prochaska, J. O., Velicer, W. F., DiClemente, C. C , Guadagnoli, E., & Rossi, J. S. (1991). Patterns of change: Dynamic typology applied to smoking cessation. Multivariate Behavioral Research, 26, 83—107. Rakowski, W., Dube, C. E., Marcus, B. H., Prochaska, J. O., Velicer, W. F., & Abrams, D. B. (1992). Assessing elements of women's decisions about mammography. Health Psychology, 11, 111-118. Rossi, J. S. (1989a). Exploring behavioral approaches to UV risk reduction. In A. Moshell & L. W. Blankenbaker (Eds.), Sunlight, ultraviolet radiation and the skin (pp. 91-93). Bethesda, MD: National Institutes of Health. Rossi, J. S. (1989b). The hazards of sunlight: A report on the Consensus Development Conference on Sunlight, Ultraviolet Radiation, and the Skin. Health Psychologist, 11(3), 4-6. Sallis, J. F., & Hovell, M. F. (1990). Determinants of exercise behavior. In J. O. Holloszy & K. B. Pandolf (Eds), Exercise and sport sciences review 18 (pp. 307-330). Baltimore, MD: Williams & Wilkins. Sonstroem, R. J. (1988). Psychological models. In R. Dishman (Ed.), Exercise adherence (pp. 125-154). City, IL: Human Kinetics. Velicer, W. F., DiClemente, C. C , Prochaska, J., & Brandenburg, N. (1985). A decisional balance measure for assessing and predicting smoking status. Journal of Personality and Social Psychology, 48, 1279—1289.
Assessing Motivational Readiness and Decision Making for Exercise Bess H. Marcus The Miriam Hospital and Brown University School of Medicine
William Rakowski Department of Community Health Center for Gerontology and Health Care Research Brown University
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Joseph S. Rossi Cancer Prevention Research Center University of Rhode Island Motivational and cognitive processes of behavior change with respect to the area of exercise adoption were investigated. A total of 778 men and women, recruited from four worksites, answered a 40-item questionnaire consisting of statements based on constructs from the transtheoretical model of behavior change. Principal-components analysis identified two factors—one a 6-item component representing avoidance of exercise (Cons), the other a 10-item component representing positive perceptions of exercise (Pros). Analysis of variance showed that the Pros, Cons, and a Decisional Balance measure (Pros minus Cons) were significantly associated with stage of exercise adoption. Results are consistent with applications of the model to smoking cessation and other areas of behavior change. Distinctions between exercise adoption and behaviors such as smoking cessation, weight loss, and alcoholism are discussed. Key words: exercise, decisional balance, health behavior, transtheoretical model, behavioral adoption
The benefits of exercise influence a wide variety of diseases and conditions affecting both physical and mental health. Regular exercise can help prevent and treat coronary heart disease, osteoporosis, diabetes, hypertension, and depression (Harris & Associates, 1989). Because the benefits of exercise are clear, researchers and clinicians are faced with two main challenges: first, how to get people to initiate exercise behavior and, second, how to help active people maintain their exercise behavior. Little is known about how to intervene in a population so as to increase and then sustain the proportion of individuals interested in adopting regular exercise. More is known about relapse after a program has been started. Research has documented that, across a variety of populations, approximately 50% of individuals who join an exercise program will drop out during the first 3 to 6 months (Carmody, Senner, Manilow, & Matarazzo, 1980; Dishman, 1988b). Although many studies have been conducted to address the problem of exercise relapse, and some have shown positive effects on shortterm adherence to relatively brief exercise programs, little success has been attained in improving long-term maintenance of exercise behavior (Dishman, 1982,1988a, 1988b; Martin & Dubbert, 1982, 1984). This pattern of exercise relapse is similar to the negatively accelerated relapse curve often seen in the addictions (Hunt, Barnett, & Branch, 1971). This study examined the potential for extending a theory of behavior change based on cognitive and social learning constructs to the area of exercise adoption. The transtheoretical model was proposed initially by Prochaska and DiClemente (1982,1983) as a general model of intentional behavior change. The core of the model is a sequence of stages along a continuum of behavioral change from
Requests for reprints should be sent to Bess H. Marcus, Division of Behavioral Medicine, RISE Building, The Miriam Hospital and Brown University School of Medicine, 164 Summit Avenue, Providence, RI 02906.
precontemplation (not intending to make changes), to contemplation (considering a change), to preparation (making small changes), to action (actively engaging in the new behavior), to maintenance (sustaining the change over time). There is also the possibility of relapsing back to an earlier stage, as change is hypothesized to occur in a cyclical rather than linear pattern. The model also defines a set of outcome or intermediate variables that includes decisional balance—the pros and cons of behavior change (Velicer, DiClemente, Prochaska, & Brandenburg, 1985). Decisional balance is based on the theoretical model of decision making developed by Janis and Mann (1977) and was initially applied to smoking cessation by Velicer et al. (1985). It is based on a comparison of the perceived positive aspects (pros) and negative aspects (cons) of a new behavior. DiClemente et al. (1991) described the difference on each of these measures across the stages of change for smoking cessation. The stages of smoking behavior have been found to correspond to differences in the balance of the perceived pros and cons of that behavior on the decisional balance measure. Persons in action and maintenance have a decisional balance favoring the positive features (pros), persons in precontemplation have a balance reflecting reasons not to change (cons), and persons in contemplation tend to fall between those in precontemplation and action. In longitudinal studies of smoking cessation, decisional balance has been especially useful in predicting movement from the precontemplation to the contemplation stage and in predicting behavior change (Prochaska, Velicer, DiClemente, Guadagnoli, & Rossi, 1991). The transtheoretical model has been applied most thoroughly to smoking cessation (DiClemente et al., 1991; Prochaska & DiClemente, 1983; Prochaska, Velicer, DiClemente, & Fava, 1988). However, it has more recently been extended to other health-related behaviors including exercise (Marcus, Banspach et al., in press; Marcus & Owen, 1992; Marcus, Rossi, Selby, Niaura, & Abrams, in press; Marcus, Selby, Niaura, & Rossi, 1992; Marcus & Simkin,
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
258
MARCUS, RAKOWSKI, ROSSI
in press), alcohol use (DiClemente & Hughes, 1990), weight loss (O'Connell & Velicer, 1988), screening mammography (Rakowski et al., 1992), and ultraviolet light exposure (Rossi, 1989a, 1989b) and may also be helpful for understanding stability and change at the level of organizations (Abrams, Emmons, Niaura, Goldstein, & Sherman, 1991). A strength of this model is its focus on the dynamic nature of health behavior change. Because this is a dynamic model, the different transitions in adoption and maintenance of a behavior described by other researchers (Dishman, 1982; Sallis & Hovell, 1990; Sonstroem, 1988) can be specifically examined. This model suggests that behavior change is not an all-or-none phenomenon and that individuals who stop performing a behavior may have intentions to start again (Sonstroem, 1988). The present article reviews questionnaire development pertinent to the decisional balance aspect of the transtheoretical model for extension of the model to exercise adoption. Our particular interests were in whether scales representing pros and cons could be developed (measurement development) and whether these scales would be associated with stages of exercise adoption (model testing) as a basis for further model development in this area of health promotion and disease prevention. METHOD
Subjects Subjects were recruited as part of a worksite health promotion project emphasizing health-risk appraisal and smoking cessation. Seven hundred seventy-eight of a possible 1,173 (66%) male and female employees of four worksites—a retail outlet, an industrial manufacturer, a government agency, and a medical center—participated. Fifty-four percent of the sample was female, average age was 41.5 years (SD = 11.0 years), and average years of education was 13.5 (SD = 2.0 years). Ninety-five percent of the subjects were White, 70% were married, and 17% were currently smoking. Median annual household income was $35,000, with 70% of employees involved in white-collar occupations. Procedure The present data were collected in the course of normal follow-up surveying at the worksites. Employees received a cover letter from their chief executive officer explaining that they were invited to continue their participation in a study about health promotion at the worksite. Persons who volunteered for the study completed questionnaires on exercise and basic demographic information and had previously provided informed consent. Respondents also completed questionnaires about their smoking status and about other lifestyle behaviors as part of a large study on health-risk status and smoking behavior at the workplace. Details concerning the larger study will be reported elsewhere. Subjects were informed that their names would be entered into a drawing for a $100 prize in return for their participation. Measures Decisional Balance. Development of the Decisional Balance measure for exercise followed the sequential method of scale construction described by Jackson (1970,1971; see also Comrey, 1988)
and involving a sequence of steps to ensure content and internal validity. An initial pool of approximately 75 statements reflecting the positive (pros) and negative (cons) aspects of exercise was generated by a small group of male and female regular exercisers and non-exercisers, including researchers and laypersons. The statements were reviewed and revised to improve clarity of expression, eliminate redundancies, and ensure representation among the decisional categories recommended by Janis and Mann (1977): (a) gains or losses expected for oneself, (b) gains or losses expected for significant others, (c) self-approval or -disapproval due to the behavior, and (d) approval or disapproval by others due to the behavior. An example of a pro item was "I would feel healthier if I exercised regularly", and an example of a con item was "I do not like the way my body feels when I exercise." The final pool of items consisted of 40 statements, including 20 pros and 20 cons of exercising. Subjects were asked to indicate, on a 5-point Likert scale ranging from not at all important (1) to extremely important (5), how important each statement was with respect to their decision to exercise or not to exercise.
Stages of adoption for exercise. Stages of adopting exercise were measured using an 11-point scale in the shape of a ladder. Each rung had a number (0 through 10), and 5 rungs also had written labels to serve as anchor points. The 0 rung was labeled / currently do not exercise and I do not intend to start exercising in the next 6 months (precontemplation), 2 was labeled / currently do not exercise, but I am thinking about starting to exercise in the next 6 months (contemplation), 5 was labeled / currently exercise some, but not regularly (preparation), 8 was labeled I currently exercise regularly but I have only begun doing so within the last 6 months (action), and 10 was labeled / currently exercise regularly and have done so for longer than 6 months (maintenance). The formal stage-of-adoption labels (just noted in parentheses) were not placed on the ladder. Respondents picked the rung on the ladder that most accurately described their current exercise behavior. Regular exercise was defined as exercising three or more times per week for at least 20 min each time (American College of Sports Medicine, 1990). Respondents could indicate numbers other than those that were labeled, but the labels corresponded to the minimum requirements for membership in a particular stage. Thus, a subject responding with a "4" on the ladder was classified as a contemplator because the minimum requirements for the preparation stage had not been met. In practice, almost 80% of respondents selected one of the five labeled response categories, suggesting that the stage ladder could be shortened in future studies. Reliability of the stages-of-exercise-adoption measure has been examined. The Kappa index of reliability over a 2-week period was .78 (N = 20; Marcus et al., 1992).1 Concurrent validity for this measure has been demonstrated by its significant association with the Seven Day Recall Physical Activity Questionnaire (Marcus & Simkin, in press). RESULTS Principal-Components Analysis Seventeen items (3 Pros, 14 Cons) from the questionnaire were eliminated due to excessive positive or negative response, which 1
Fleiss (1981) reported that values of Kappa above .75 indicate strong agreement.
259
ADOPTION OF EXERCISE
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
limited the variance of the items. The remaining 23 items were subjected to a principal-components analysis with a varimax rotation. The number of components to be retained was determined by the scree method (Cattell, 1966) and interpretability. After rotation, an item was judged to be salient for a component if there was a component loading of .50 or greater and if the item did not load on another component. TABLE 1 Results of Principal-Components Analysis for Statements Representing Positive and Negative Orientations Toward Exercise (Pros and Cons) Component Loading
Statement Wording Pros I would have more energy for my family and friends if I exercised regularly. Regular exercise would help me relieve tension. I would feel more confident if I exercised regularly. I would sleep more soundly if I exercised regularly. I would feel good about myself if I kept my commitment to exercise regularly. I would like my body better if I exercised regularly. It would be easier for me to perform routine physical tasks if I exercised regularly. I would feel less stressed if I exercised regularly. I would feel more comfortable with my body if I exercised regularly. Regular exercise would help me have a more positive outlook on life. Cons I think I would be too tired to do my daily work after exercising. I would find it difficult to find an exercise activity that I enjoy that is not affected by bad weather. I feel uncomfortable when I exercise because I get out of breath and my heart beats very fast. Regular exercise would take too much of my time. I would have less time for my family and friends if I exercised regularly. At the end of the day, I am too exhausted to exercise.
.77 .81 .86
Two components were interpreted from the varimax pattern. The first component was composed of 6 items that highlighted unfavorable perceptions of exercise and reasons not to exercise. Conceptually, this factor represented the Cons dimension evident in previous work with the transtheoretical model. The second factor was composed of 10 items. This second category represented a positive or pro perception of exercise. The 6 Cons and 10 Pros items were subjected to another principal-components analysis in order to determine their final loadings on their respective components. The two components accounted for 60.4% of the total item variance. The loadings for each of the 16 items are provided in Table 1. Each scale was replicated in the second analysis (i.e., all loadings were greater than .50, and no other components emerged). Coefficient alpha (internal consistency) reliabilities were satisfactory (Cons = .79, Pros = .95). Stages of Adoption
.72 .86 .81 .84 .83 .85 .85
.71 .52
.62
.79 .70 .78
One-way analysis of variance was used to examine the association among stages of exercise adoption and the Pros and Cons indices. In order to provide a standard metric, the Pros and Cons indices were converted to T scores (M = 50, SD = 10). In addition, the Decisional Balance measure was created by subtracting Cons from Pros. Table 2 presents the T-score means and standard deviations by stage of exercise adoption. Differences on the Pros scale by stage of adoption were significant, F(4, 712) = 34.84, p < .0001, t]2 = .16. Based on a NewmanKeuls analysis, Pros scores were significantly higher for subjects in maintenance compared to precontemplation and contemplation; action compared to precontemplation, contemplation, and preparation; and preparation and contemplation compared to precontemplation. Therefore, 7 of the 10 possible pairwise contrasts were significant. Differences on the Cons scale by stage of adoption were also significant, F(4, 710) = 25.31, p < .0001, T)2 = .12. Based on a Newman-Keuls analysis, Cons scores were significantly lower for subjects in maintenance compared to action; action and maintenance compared to preparation; maintenance compared to precontemplation; and all stages compared to contemplation. Thus, 8 of the 10 possible pairwise contrasts were significant. Last, the Decisional Balance analysis also achieved significance, F(4, 709) = 49.32, p < .0001, v\2 = .22. Newman-Keuls follow-up indicated that all stages were significantly different from all other stages, with precontemplators scoring lowest and maintainers scoring highest. Figure 1 presents a pictorial view of the Pros and Cons scales by stage of exercise adoption.
TABLE 2 Means and Standard Deviations of the Pros, Cons, and Decisional Balance Scales by Stage of Exercise Adoption Stage of Adoption Precontemplation^ Scale
M
Pros Cons Decisional Balance
36.75 50.46
a
SD
c
11.13 d
e
Preparation0
Action6
Maintenance"
SD
M
SD
M
SD
M
SD
49.56 53.79
8.95 9.25
50.72 50.32
8.78 9.01
52.97 47.97
7.66 9.44
53.02 44.10
9.89 8.49
-4.21
11.16
0.40
11.97
5.00
12.21
9.09
13.12
M
11.33 11.93
-13.73 b
Contemplation^
n = 53. n = 242. n = 182. n = 101. n = 139.
260
MARCUS, RAKOWSKI, ROSSI
60
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
50
40
30
PreContemplation
Cuntemplation
Preparation
Action
Maintenance
FIGURE 1 T-score means for the Pros and Cons scales by stage of exercise adoption.
DISCUSSION This investigation provided encouraging results for extending the transtheoretical model of behavior change to the area of exercise adoption. Three important objectives were supported. First, a pool of items was developed guided by the model and yielded principal components consistent with major decisional balance constructs. This outcome was important for measurement development. Second, the decisional balance measures reflected hypothesized differences across stages of exercise adoption. This outcome was important for model testing. Last, and due to these two points, there is reason to pursue additional model development with exercise. The generalizability of these findings is limited in that they are based on a single cross-sectional investigation utilizing self-report data with no objective information on actual exercise behavior. However, the present results do provide some evidence that participants' reports of current exercise behavior (i.e., stages of adoption) correspond to beliefs about favorable and unfavorable features (i.e., pros, cons, decisional balance). Conversely, knowing participants' beliefs may portend the degree of acceptance or reluctance encountered by attempts to produce behavioral change toward regular exercise behavior. That the transtheoretical model appears to translate to exercise is encouraging, because exercise adoption involves the initiation of a positive behavior for health promotion and disease prevention in a general population rather than the elimination of a potentially harmful behavior like smoking or the reduction of a clinically assessed condition (e.g., obesity).
The 16 items eventually used from the present instrument provided indices with adequate empirical properties and conceptual fit to the model. At the same time, the items were only a subset of our pool, and there is certainly ample room for other investigators to test instruments of their own design. Replication of the pattern of statistical significance found here, but using other instruments, could add further support for the extension of the model to exercise. Additional investigations examining the reliability and validity of these measures are warranted. Because intervention is the ultimate objective of the model, an important consideration is that there be strong evidence in favor of the efficacy of the newly targeted health behavior. The consensus for the benefits of smoking cessation and exercise adoption make them appropriate foci for the model. If the present cross-sectional findings hold up in longitudinal studies of exercise behavior, it will be essential for interventions to be aimed at increasing participants' pros of exercise in order for movement through the stages to be accomplished. Although helping participants to decrease their cons of exercise is also important, this alone is not likely to result in enhanced readiness to be active. Knowledge of participants' pro and con beliefs in addition to the sociodemographic, biologic, and environmental influences on them (Dishman, 1982; Martin & Dubbert, 1982) will enhance the ability of researchers and clinicians to design effective exercise promotion interventions. An exploration of specific Pros and Cons items reveals many themes in common with previous empirical observations regarding the determinants of exercise. Pros items focusing on enhanced
ADOPTION OF EXERCISE confidence, feeling good about oneself, and having more energy for one's family and friends support the determinants-related work of Dishman, Sallis, and Orenstein (1985) and Sallis and Hovell (1990). Cons items such as being too tired to exercise, being concerned about bad weather, feeling uncomfortable and out of breath, and not having enough time also support previous research on the determinants of exercise. The transtheoretical model provides a framework for integrating these diverse variables and creating an empirical index that can be used to better understand the determinants of exercise and therefore to design more efficacious exercise interventions.
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ACKNOWLEDGMENTS This research was partially supported by National Institutes of Health Grant SO7RR05818 and National Cancer Institute Grant CA50087. We thank Elaine Taylor, Sheilarae Carpentier, Anna Klevak, Nancy Keenan, and Mark Morgenstern for their efforts on this project.
REFERENCES Abrams, D. B., Emmons, K. M., Niaura, R. S., Goldstein, M. G., & Sherman, C. B. (1991). Tobacco dependence. In P. E. Nathan, J. W. Langenbucher, B. S. McCrady, & W. Frankenstein (Eds.), The annual review of addictions treatment and research (Vol. 1, pp. 391—436). New York: Pergamon. American College of Sports Medicine. (1990). Position statement on the recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness in healthy adults. Medicine and Science in Sports and Exercise, 22, 265-274. Carmody, T. P., Senner, J. W., Manilow, M. R., & Matarazzo, J. D. (1980). Physical exercise rehabilitation: Long-term dropout rate in cardiac patients. Journal of Behavioral Medicine, 3, 163—168. Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276. Comrey, A. L. (1988). Factor-analytic methods of scale development in personality and clinical psychology. Journal of Consulting and Clinical Psychology, 56, 754-761. DiClemente, C. C, & Hughes, S. (1990). Stages of change profiles in outpatient alcoholism treatment. Journal ofSubstance Abuse, 2,217-235. DiClemente, C. C , Prochaska, J. O., Fairhurst, S., Velicer, W. F., Velasquez, M., & Rossi, J. (1991). The process of smoking cessation: An analysis of precontemplation, contemplation, and preparation stages of change. Journal of Consulting and Clinical Psychology, 59, 295-304. Dishman, R. K. (1982). Compliance/adherence in health-related exercise. Health Psychology, 1, 237-267. Dishman, R. K. (1988a). Exercise adherence research: Future directions. American Joural of Health Promotion, 3, 52—56. Dishman, R. K. (1988b). Overview. In R. Dishman (Ed.), Exercise adherence (pp. 1-9). City, IL: Human Kinetics. Dishman, R. K., Sallis, J. F., & Orenstein, D. R. (1985). The determinants of physical activity and exercise. Public Health Reports, 100, 158-172. Fleiss, J. L. (1981). Statistical methods for rates and proportions (2nd ed.). New York: Wiley. Harris, L, & Associates, Inc. (1989). The prevention index '89: summary report. Emmaus, PA: Rodale. Hunt, W. A., Barnett, L. W., & Branch, L. G. (1971). Relapse rates in addictions programs. Journal of Clinical Psychology, 27, 455-456.
261
Jackson, D. N. (1970). A sequential system for personality scale development. In C. D. Spielberger (Ed.), Current topics in clinical and community psychology (Vol. 2, pp. 61-96). New York: Academic. Jackson, D. N. (1971). The dynamics of structured personality tests. Psychological Review, 78, 229-248. Janis, I. L., & Mann, L. (1977). Decision making: A psychological analysis of conflict, choice and commitment. New York: Free Press. Marcus, B. H., Banspach, S. W., Lefebvre, R. L., Rossi, J. S., Carleton, R. A., & Abrams, D. B. (in press). Using the stages of change model to increase the adoption of physical activity among community participants. American Journal of Health Promotion. Marcus, B. H., & Owen, N. (1992). Motivational readiness, self-efficacy and decision-making for exercise. Journal ofApplied Social Psychology, 22, 3-16. Marcus, B. H., Rossi, J. S., Selby, V. C , Niaura, R. S., & Abrams, D. B. (in press). The stages and processes of exercise adoption and maintenance in a worksite sample. Health Psychology. Marcus, B. H., Selby, V. C , Niaura, R. S., & Rossi, J. S. (1992). Self-efficacy and the stages of exercise behavior change. Research Quarterly for Exercise and Sport, 63, 60—66. Marcus, B. H., & Simkin, L. R. (in press). The stages of exercise behavior. Journal of Sports Medicine and Physical Fitness. Martin, J. E., & Dubbert, P. M. (1982). Exercise applications and promotion in behavioral medicine: Current status and future directions. Journal of Consulting and Clinical Psychology, 50, 1004-1017. Martin, J. E., & Dubbert, P. M. (1984). Behavioral management strategies for improving health and fitness. Journal of Cardiac Rehabilitation, 4, 200-208. O'Connell, D., & Velicer, W. F. (1988). A decisional balance measure for weight loss. International Journal of Addictions, 23, 729—750. Prochaska, J. O., & DiClemente, C. C. (1982). Transtheoretical therapy: Toward a more integrative model of change. Psychotherapy: Theory, Research, andPractice, 20, 161-173. Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self change of smoking: Toward an integrative model. Journal of Consulting and Clinical Psychology, 51, 390-395. Prochaska, J. O., & DiClemente, C. C. (1985). Common processes of self-change in smoking, weight control, and psychological distress. In S. Shiftman & T. Wills (Eds.), Coping and substance use (pp. 345-363). New York: Academic. Prochaska, J. O., Velicer, W. F., DiClemente, C. C , & Fava, J. (1988). Measuring processes of change: Applications to the cessation of smoking. Journal of Consulting and Clinical Psychology, 56, 520—528. Prochaska, J. O., Velicer, W. F., DiClemente, C. C , Guadagnoli, E., & Rossi, J. S. (1991). Patterns of change: Dynamic typology applied to smoking cessation. Multivariate Behavioral Research, 26, 83—107. Rakowski, W., Dube, C. E., Marcus, B. H., Prochaska, J. O., Velicer, W. F., & Abrams, D. B. (1992). Assessing elements of women's decisions about mammography. Health Psychology, 11, 111-118. Rossi, J. S. (1989a). Exploring behavioral approaches to UV risk reduction. In A. Moshell & L. W. Blankenbaker (Eds.), Sunlight, ultraviolet radiation and the skin (pp. 91-93). Bethesda, MD: National Institutes of Health. Rossi, J. S. (1989b). The hazards of sunlight: A report on the Consensus Development Conference on Sunlight, Ultraviolet Radiation, and the Skin. Health Psychologist, 11(3), 4-6. Sallis, J. F., & Hovell, M. F. (1990). Determinants of exercise behavior. In J. O. Holloszy & K. B. Pandolf (Eds), Exercise and sport sciences review 18 (pp. 307-330). Baltimore, MD: Williams & Wilkins. Sonstroem, R. J. (1988). Psychological models. In R. Dishman (Ed.), Exercise adherence (pp. 125-154). City, IL: Human Kinetics. Velicer, W. F., DiClemente, C. C , Prochaska, J., & Brandenburg, N. (1985). A decisional balance measure for assessing and predicting smoking status. Journal of Personality and Social Psychology, 48, 1279—1289.
