Pr/chologicol Reporh, 1991, 68, 1327-1335. @ Psychological Reports 1991

COGNITIVE STRATEGIES AND PERSONALITY VARIABLES IN ADHERENCE TO EXERCISE '

University of South Alabama Summary.-Variables associated with adherence to a 6-wk. jogging program were examined in a group of 26 sedentary women, who received insrruct~onand redorcement for their efforts: one group of 16 received instruction in cognitive self-statements; the other group of 10 did not. Measures included a battery of psychological questionnaires, percentage of body fat, weight, height, and distance run in a 12-min. test. Subjects maintained logs in which they recorded Frequency and duration of exercise. Follow-up data were collected at 6-wk. and 6-mo. postprogram. Initial compliance was related to scores on the Self-motivation Inventory, the Type H (Hard-driving) scale of the Jenkins Activity Survey, and the State Anxiety Scale. At 6-mo. follow-up, none of the measures were significantly correlated with exercise frequency.

The beneficial effects of regular aerobic exercise have been well documented and include physical benefits such as reduced miglyceride and LDL cholesterol levels, improved cardiopulmonary functioning, control of obesity, and better management of hypertension and diabetes (Martin & Dubbert, 1982, 1987). Paffenbarger, Wing, and Hyde (1978) indicate that individuals who engage in moderate physical activity throughout their lives live longer than comparable individuals who get little exercise. In addition, beneficial psychological effects related to exercise have also been reported, including decreases in depression and anxiety and increases in self-esteem (Folkins & Sime, 1981; LabbC, Welsh, & Delaney, 1988; Morgan & O'Conner, 1988). Despite these positive benefits of aerobic exercise, adherence to a regular exercise regimen is typically poor. Recent surveys indicate that two-thirds of Americans do not exercise regularly and 45% may not exercise at all (Martin & Dubbert, 1987; Dishman, Sallis, & Orenstein, 1985). Studies indicate that of those who begin an exercise program, whether on their own or in a structured program, roughly one-half or fewer will still be exercising after three months (Martin & Dubbert, 1982). This decline in adherence to exercise is similar in studies of individuals with known health problems, such as inlviduals in a cardiac rehabilitation program (Oldridge, 1988) as well as in studies of healthy adult subjects (Martin & Dubbert, 1982). Only recently have researchers begun to examine the factors associated with adherence to exercise. Interestingly, neither one's attitude nor one's awareness of the benefits of exercise predicts participation in or later adherence to an exercise program (Dishman, et al., 1985). Everybody seems to 'Address correspondence to M. Cay Welsh, Ph.D., Department of Psychology, University of South Alabama, Mobile, AL 36688.

1328

M. C. WELSH, ETAL.

endorse the idea that exercise is good for them but relatively few engage in it. However, Dishman (1982) did find that low scores on the self-motivation factor of h s Self-motivation Inventory did predict attrition from exercise. Dishman and Gettman (1980) have proposed a psychobiologic model of exercise compliance based on inventory scores and measures of body composition. They claim these variables can correctly identify 80% of individuals as potential adherers or dropouts. They have found that individuals with a high percentage of body fat are less likely to adhere to an exercise program. I n a series of six studies on exercise adherence (Martin, Dubbert, Katell, Thompson, Raczynslu, Lake, Smith, Webster, Sikora, & Cohen, 1984) several behavioral techniques such as contracting and a lottery procedure to reinforce compliance improved short-term compliance. However, these techniques were not effective in maintaining long-term adherence. Individual praise and feedback during exercise resulted in better attendance and adherence than did group praise. Similarly, joggers who set flexible, distal goals were more likely to adhere to an exercise program than individuals who were encouraged to set more rigid, immediate goals. Martin, et al. (1984) proposed that cognitive strategies may also be important in exercise adherence. They instructed one group of beginning joggers to use cognitive association strategies which focused on awareness of internal sensations while exercising. These strategies were described as being similar to traditional coaching strategies. For example, these subjects were encouraged to set high personal goals for themselves and to tell themselves statements such as "do better than last time, you can run furtherlfaster." The associative strategies were designed to focus on internal stimuli. Such cognitive statements and the use of self-monitoring have been reported to be used by elite runners during races. Another group of subjects was instructed to use cognitive dissociative strategies, which involved focusing on external stimuli during exercise. These external stimuli were thought to be pleasant and distracting from the discomfort of exercise, for example, "smell the flowers" rather than focus on potentially unpleasant bodily sensations. The group using dissociative strategies had a significantly higher compliance rate both during the program and adherence at 6-mo. follow-up. While these researchers described the two strategies as differing on the dimension of internal vs external awareness during exercise, another potential difference was the competitiveness associated with the two strategies. The associative strategies encouraged setting more stringent goals, i.e., running harder, while the dissociative strategy seemed to make no reference to intensity of effort. It may be that with beginning joggers, a strategy which emphasizes competitiveness or quick improvement sets too high a standard. This explanation is consistent with Bandura's (1982) concept of self-efficacy, in that beginning joggers may not perceive themselves as capable of performing in a physically

PERSONALITY I N ADHERENCE TO EXERCISE

1329

competitive or highly competent manner. Associative cognitive strategies focusing on body awareness or running as an activity could be used and could be worded as noncompetitive. This change in focus may be as effective for beginning exercisers as dissociative strategies. The influence of personality variables on exercise adherence has not been systematically studied. For example, one study (Rejeslu, Morley, & Miller, 1984) shows Type A behavior pattern to be positively related to fitness gains, while another indicates that Type A behavior is highly associated with early dropouts from a cardiac rehabhtation exercise program as well as withdrawal from participation in corporate exercise programs (Oldridge, 1982; Shephard & Cox, 1980). Similarly, there have been some indications that depressive symptoms have been negatively associated with amount of activity, but there has been little direct examination of the effects of such variables as depression or anxiety on adherence to a structured program (Dishman, et al., 1985). Similarly, evidence has been contradictory on the importance of health beliefs and exercise (Dishman, et al., 1985). The current study was conducted to examine variables which may be associated with compliance to a jogging program and adherence to continued aerobic exercise. Compliance refers to following the directive during the structured program of exercising a minimum of three times weekly, for 20 to 30 minutes within 60 to 80% of maximum heart rate. Adherence refers to the length of time the individual continues with the program regimen. It was hypothesized that individuals who received instruction in cognitive strategies would comply better with the program and would adhere to an exercise regimen longer than subjects who were given no instruction in cognitive strategies. As a result of compliance to the program women would improve their running ability by the end of the 6-wk.period. It was hypothesized that the women who complied with the 3-day exercise regimen would be more likely to adhere to exercise after the end of the structured program. It was further hypothesized that adherence would be associated with amount of body fat, scores on the Self-motivation Inventory, the Beck Depression Inventory, Health Locus of Control, and Jenkins Activity Survey.

Subjects Twenty-six women (M age = 35.7 yr., range of 20 to 50 yr.) were recruited from a community to begin a jogging program. Most of the women reportedly applied to a study on weight control and 60% of the subjects were at least 20% overweight (M weight = 163.5 Ib., range 129 to 220 pounds; M body mass index = 26.6, range 22 to 34). Thls is a higher mean weight than reported in most exercise studies and, in fact, approaches the weights of individuals in weight-loss studies. None had been involved in a

1330

M. C. WELSH, ETAL.

consistent aerobic exercise program within the previous year. All subjects obtained medical consent to participate. Measures State-Trait Anxiety Inventories (STAI Form X).-The two scales, state and trait anxiety, consist of 20 items each that the respondent rates on a scale of 1 (not at all) to 4 (very much so). On the state scale the individual rates how she feels at the present; on the trait scale the rating is for how the individual generally feels. Test-retest reliabilities for the Trait-Anxiety scales for male and female undergraduates over 6 mo. ranged from .73 to .77, respectively. Internal consistency of the Anxiety-State scales ranges from .83 to .92 (Spielberger, 1985). Beck Depression Inventory.-This inventory has 21 items which include four statements ranked in order of severity of experience of a depressive symptom. The subject is to choose the statement which most closely reflects her present state. Each alternative is assigned a nonempirical weight of 0 to 3. Internal consistency is indicated by odd-even item correlation of .86 (Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961). Test-retest reliability of .74 over a 3-mo. interval was reported for undergraduate subjects (Miller & Seligman, 1973). Multidimensional-he& Locus of Control.-This scale has 18 items which measure three dimensions: Internality, Chance Externahty, and Powerful Others Externality, each consisting of 6-item scales with a 6-point Likert format. Reported alpha reliabilities ranged from .83 to .86 (Wallston & Wallston, 1981). Jenkins Activity Suruqr.-The survey contains 52 multiple-choice questions designed to measure the Type A behavior pattern thought to be strongly associated with the risk of coronary heart disease. Two of the possible four scales were used, Type A and Type H (hard-driving). The subscales have reliability coefficients ranging from .73 to .85. Many validity studies have indicated clear agreement between the survey and the structured interview (Jenkins, Zyzanski, & Rosenman, 1979). Self-motivation Inventory.-This inventory contains 40 items which are rated on a 5-point scale to indicate that the items are extremely uncharacteristic to extremely characteristic of the individual. The items pertain to general goal-striving, persistence, and will power. Dishman and Ickes (1981) report h g h stability (rs ranged from .86 to .92, with repeated measures taken from 1 to 5 months). Procedure Two exercise times were made available to subjects and the women signed up for their preferred meeting times. Groups were then randomly assigned to one of two conditions: Group 1 (n = 10) received exercise instruc-

PERSONALITY I N ADHERENCE TO EXERCISE

133 1

tions and verbal reinforcement for their physical activity, but no instruction on cognitive strategies. Group 2 (n = 16) received the same exercise instruction and reinforcement but were also taught positive cognitive selfstatements related to exercise. The initial meeting was spent taking prerneasures on the questionnaires and physical measures of height, weight, and percent body fat. Body fat was estimated by measuring subcutaneous skinfold fat with a Lange caliper, and measurements were taken by a trained technician. Skinfold sites measured included triceps, iliac, and thigh. Measurement of distance run in a 12-min. interval on a %-mile outdoor rubberized track was also taken. Subjects met in groups twice weekly with instructors over a 6-wk, training period. Class sessions included approximately 20 rnin. of instruction about jogging, followed by warm-up and stretching and approximately 30 min. of walking and jogging on the ]A -mile outdoor track. I n addition to the scheduled group meetings, the women were instructed to jog on their own for 20 to 30 minutes at least one other day of the week. Compliance was measured by attendance in the program and the recording of exercise behavior in log books which were turned in weekly. Subjects recorded their running activity both for the class times and for jogging on their own. During the structured sessions, instructors jogged with the subjects, modeling appropriate procedures and providing reinforcement for subjects' efforts. The women were instructed in calculation of target heart rate at 60 to 80% of their age-predicted maximum heart rate and shown how to take their own pulse. Subjects were instructed to check periodically and record their pulse during exercise and at the end of the exercise period, with the goal of staying within the target heart-rate range for 20 to 30 min. during each session of exercise. For the cognitive self-statement group, they were instructed in sample self-statements as well as had examples printed in the daily log book in which they recorded their exercise behavior. Examples of self-statements were "notice your rhythm and smoothness. Think of moving lightly over the ground. Think smooth and steady," or "notice your environment, the grass, the clouds, etc." For other subjects, the self-statements were omitted from the log books. Subjects maintained the daily logs in which they recorded frequency and duration of exercise, postexercise heart rate, and for the self-statement group, the subjects' use of positive self-statements. The final session included the postassessment on psychological and physical measures. The women were contacted 6 weeks and 6 months later for follow-up to assess whether they continued to exercise and, if so, the average miles (or time) and frequency per week.

RESULTS Of the 26 subjects who began the study, one subject from Group 1 dropped out and three subjects from Group 2 dropped out during the pro-

1332

M. C. WELSH, ETAL

gram. At 6 weeks follow-up, eight of the remaining 22 subjects reported they no longer exercised three or more days per week; however, 64% of subjects reported continuing regular exercise. At 6 months follow-up, 17 of the subjects could be contacted and 11 of those indicated regular aerobic exercise. This again indicates approximately 64% exercising. A training effect was indicated by the improvement in performance on the 12-min. run test. A t test was performed on distance run at the pre- and post-6-wk. training period (M pretraining = 4308.4 ft., SD = 602.0; M posttraining = 5147.1 ft., SD = 805.6, p < .001). All subjects showed some improvement in distance run in the 12-min. interval; the range of improvement was 1 to > I % , with mean improvement at 19.8%. See Table 1 for the descriptive statistics on the variables measured. TABLE 1 DESCRIPTIVESTATISTICSFORMEAsLJRED V ~ L E S Variable

N

M

SD

Age (yr.) Body Mass Index Weight (Ib.) Self-motivation Type A Type H Beck-Pretraining State Anxiety-Pretraining Trait Anxiety-Pretraining Locus of Control-Pretraining Attendance Beck-Posttraining State Anxiety-Posttraining Trait Anxiety-Posttraining Locus of Control-Posttraining Distance-Pretraining Distance-Posttraining Milesweek

To evaluate the effect of the cognitive self-statements 2 x 2 analyses of variance with repeated measures were performed. These analyses yielded no significant group differences on distance run, compliance, or any personality variables. Pearson product-moment correlations were calculated on the physical measures, exercise frequencies, attendance to exercise class, and scores on the personality measures. Distance run during both the pre- and post-12-rnin. run was negatively correlated with weight ( r = - 3 7 , p < . 0 5 and r = - . 7 7 , p < ,001) and distance run during the posttraining 12-min. run was also negatively correlated with percentage of body fat (r = -.51, p < .05); see Table 2.

PERSONALITY IN ADHERENCE TO EXERCISE TABLE 2 PEARSOACORRELATIONS WITH PHYSICAL MEASURES Distance (Pre) Distance (Post) Distance (Pre)'

1.00

Distance ( P o ~ t ) ~

.76$ 1.00

Weight % Body Fat 'Distance (in feet) run in 12-min. pretest. b~istance(in feet) run in 12-min, posttest. *p

Cognitive strategies and personality variables in adherence to exercise.

Variables associated with adherence to a 6-wk jogging program were examined in a group of 26 ++sedentary women, who received instruction and reinforce...
352KB Sizes 0 Downloads 0 Views