Estimates on the basis of the National Health and Nutrition Examination Survey, using measured heights and weights, indicate that roughly 16% of American children and adolescents aged 2-19 yr are overweight (26), placing them at increased risk for Type 2 diabetes (21) and heart disease (34). Although there is debate as to the reasons for the increasing rates of pediatric obesity, evidence indicates that insufficient participation in regular physical activity plays a contributing role (20,30). Epidemiologic data indicate that a troubling proportion of adolescents are not meeting the current health guidelines for physical activity (12,33) and, despite a growing body of research in this area, little is understood about how to reverse this trend. Throughout adolescence, there is a precipitous decline in physical activity participation (9), and sedentary adolescents are at high risk for remaining sedentary into adulthood (1,2).
Most research into why some adolescents remain physically active while others do not has been based on constructs derived from social cognitive models of health behavior, such as perceived benefits, perceived barriers, and self-efficacy. In recent years, another class of variables that has received a fair amount of attention has been dimensions of the physical environment, such as urban design. Far fewer studies have attempted to identify individual differences in personality that might contribute to adolescents' likelihood of remaining active.
Personality refers to an enduring and consistent pattern of thoughts, feelings, and actions. A meta-analysis of research examining the association between personality and physical activity (28) found that the dimension of personality that was most consistently associated with physical activity was extraversion (i.e., the tendency to be sociable, assertive, to seek excitement, and to experience positive affect). Carver et al. (7) posit that extraversion reflects the "approach" component of a dual model of personality that bisects motivation and behavior into two types of action tendencies: approach and avoidance (or withdrawal).
Two neurobehavioral systems have been proposed to control the approach and withdrawal tendencies. These systems have been termed the behavioral activation system (BAS) and the behavioral inhibition system (BIS) (8,17,35). In essence, the BAS is responsible for guiding goal-directed or engagement behavior. Activity in this system causes the individual to begin (or increase) movement toward goals and to experience positive emotions. In contrast, the BIS is responsible for guiding behavioral responses to threats and novel stimuli. Activity in this system causes the individual to withdraw or inhibit ongoing behavior and to experience negative emotions. According to this model, an individual with a highly activated BAS is relatively more sensitive to cues of impending reward and is predisposed to experience positive affect in response to an emotion-eliciting stimulus. In contrast, an individual with a highly activated BIS is relatively more sensitive to cues of impending punishment and is predisposed to experience negative affect in response to an emotion-eliciting stimulus.
Carver and White (8) have developed and validated a paper-and-pencil measure (the BIS/BAS) that assesses individual differences in approach behaviors. The items on the BIS scale assess anxiety in response to a threatening situation, whereas the items on the BAS scale tap into the degree to which an individual tends to engage in approach-motivated activity toward a goal. Most of the research using these scales has been designed to explore addictive and/or pathologic behavior (17,22) or to identify physiological markers of BIS/BAS activation (4,6). Surprisingly, considering that the BIS and BAS are theorized to relate to individuals' motivation to engage in a goal-directed behavior, very little research has examined the use of BIS/BAS for understanding health-related behaviors that require active engagement, such as participation in regular physical activity.
BIS/BAS AND PHYSICAL ACTIVITY
A review of the relevant literature revealed one study that explicitly examined the association between BIS/BAS and physical activity-related constructs. In the study by Hall et al. (18), BAS and BIS scores were negatively and positively correlated with RPE, respectively, during an exercise task. Interestingly, the correlations between BAS and RPE disappeared at higher exercise intensities, but those between BIS and RPE did not. These results suggest that individuals who scored higher on the BIS scale were attending to a greater degree to the negative sensations that accompany exercise, regardless of the exercise intensity. In contrast, individuals who scored higher on the BAS may have paid greater attention to the positive sensations that accompany low-intensity exercise, but this effect disappeared at higher levels of intensity.
The present study extends these findings by examining the hypotheses that 1) both the BIS and the BAS will be related to enjoyment of exercise, preference and tolerance for high-intensity exercise, and physiological indicators of physical activity participation (cardiovascular fitness and percent body fat); 2) adolescents with a highly activated BIS will report more negative affect in response to exercise regardless of the intensity of the exercise; and 3) adolescents with a highly activated BAS will report a more positive affective response to exercise at a moderate but not at a hard level of intensity.
Understanding individual differences in affective response to physical activity is critical to the future success of interventions designed to promote adolescent physical activity (27). Personality types that may predispose individuals to respond to physical activity with a particular affective profile, for the most part, have not been explored among adolescents. Knowing which subgroups of adolescents are least likely to possess personalities that predispose them to enjoy being physically active would enable interventionists to design targeted programs for these at-risk individuals. The present study examines the role of the BIS and BAS as correlates of physical fitness, exercise enjoyment, preference for and tolerance of high-intensity exercise, and affective response to exercise among adolescents.
One hundred and forty-six healthy adolescents (55% were boys; 66% white, 10% Asian, 16% Latino, 8% other; mean (SD) age = 14.79 (0.46) yr) were recruited from two public high schools in Southern California. To be eligible for the study, adolescents had to meet the following criteria: 1) no health problems that precluded participating in regular moderate to vigorous exercise; 2) enrolled in 10th grade during the study; 3) right-handed (determined using the Chapman Handedness Scale ); 4) no history of neurological disorders, stroke, or significant head trauma (i.e., head trauma associated with loss of consciousness for >24 h); 5) not clinically depressed (defined as ever having been diagnosed with clinical depression or scoring 20 or higher on the Beck Depression Inventory ). Inclusion criteria 3), 4), and 5) were incorporated because the study also involved collection of a resting electroencephalogram (data reported elsewhere).
All study procedures were reviewed and approved by the university-based Institutional Review Board. Participants were recruited using flyers distributed at the schools. An initial telephone screening was followed by an orientation session during which potential participants completed consent forms and screening questionnaires. All participants provided written assent to participate, and consent was also obtained from a parent or guardian. Adolescents who met the study inclusion criteria then visited the university-based General Clinical Research Center, where they completed a cardiovascular fitness test (see below for procedural details). After the fitness test, each participant's ventilatory threshold (VT) was determined offline. The VT was determined by visually inspecting the ventilatory equivalents for oxygen and carbon dioxide and by identifying the point at which there was a systematic increase in ventilatory equivalent for oxygen without a corresponding increase in ventilatory equivalent for carbon dioxide. The work rate at the VT was used to calculate the target work rate for the moderate- and hard-intensity exercise tasks as described below.
A second clinic visit was scheduled within approximately 1 wk of the first. During the second visit, participants completed a computerized questionnaire that assessed BIS/BAS, enjoyment of exercise, and preference and tolerance for high-intensity exercise, and engaged in a 30-min exercise task on a cycle ergometer. Before commencing the exercise task, participants completed two affect assessments (one single-item measure-Feeling Scale (FS); and one multiple-item measure-Activation Deactivation Adjective Check List (AD ACL)). Throughout the exercise task and after a 10-min recovery period, participants were asked every 10 min to indicate their affect on the single-item measure. On conclusion of the task, at 30 min, and again after 10 min of resting recovery, participants completed the multiple-item measure of affect. Thus, the FS measure was collected at 0, 10, 20, 30, and 40 min, whereas the AD ACL was collected at 0, 30, and 40 min.
The exercise task was repeated at a third visit to the clinic, with visits 2 and 3 randomly determined to be at either a moderate-intensity level of exertion (target work rate equivalent to 80% of the watts at which the participant reached VT) or a hard-intensity level of exertion (target work rate equivalent to the average of the individual's work rates at VT and at V˙O2 peak). Exercise technicians were instructed to be alert for signs of fatigue on the part of participants and to reduce the target work rate in increments of 5-10 W after the first 10 min of exercising, as needed, to ensure that the participant would be able to complete the 30-min task. Fatigue was defined as the participant being unable to maintain a cadence of 60 rpm for 60 s. Participants were not informed as to the intensity level at which they would be exercising. Participants received monetary compensation ($25) after each clinic visit.
Each participant performed a ramp-type progressive cycle ergometer exercise test (35). Breath-to-breath measurement of gas exchange (ventilation, oxygen uptake, and carbon dioxide output) was viewed online and was subsequently analyzed using a metabolic system (Sensor Medics, Yorba Linda, CA). Data from the ramp test was used to determine the peak V˙O2 (V˙O2peak) and the VT for each participant.
Dual-energy x-ray absorptiometry
Percent body fat was determined by dual-energy x-ray absorptiometry (DEXA) using a densitometer (QDR 4500; Hologic Inc, Bedford, MA). Participants were scanned in light clothing while lying supine. DEXA scans were performed and analyzed using pediatric software. On the day of each test, the DEXA instrument was calibrated using procedures provided by the manufacturer.
Approach (BAS) and withdrawal (BIS) motivations were measured with the 20-item BIS/BAS Scale (8), which has been shown to be valid and reliable among adults, has demonstrated convergent and discriminant validity in relation to other measures of personality among college students (8) and retains its measurement and construct validity when used with adolescents (11). The 7-item BIS subscale assesses apprehension over the possibility of negative events and reaction to these events when they transpire. The 13 BAS items map onto three subscales (Drive, Reward-responsiveness, and Fun-seeking). A 5-point response scale from 1 (strongly disagree) to 5 (strongly agree) was used for items such as "If I think something unpleasant is going to happen, I usually get pretty worked up" (BIS) and "when I get something I want, I feel excited and energized" (BAS). In this sample, the BIS and BAS scales had acceptable internal consistency (Cronbach's α values ranged from 0.62 to 0.77).
Preference for and tolerance of exercise intensity
The Preference for and Tolerance of Exercise Intensity (PRETIE-Q) is a 16-item tool designed to assess the participants' preference for and tolerance of high-intensity exercise. Developed by Ekkekakis et al. (13) and validated by Ekekkakis et al. (16), there are eight items on each subscale (i.e., the Preference and Tolerance subscales), each of which is scored on a 5-point response scale ranging from 1 (I totally disagree) to 5 (I totally agree). Responses were provided to items such as "while exercising, I try to keep going even after I feel exhausted" (Tolerance) and "Exercising at a low intensity does not appeal to me at all" (Preference). In this study, the Preference and Tolerance subscales had good internal consistency (Cronbach's α = 0.85 and α = 0.82, respectively).
Enjoyment of exercise
Exercise enjoyment was assessed using the Physical Activity Enjoyment Scale (PACES) (23), as modified and validated by Motl et al. (25). The PACES contains 16 items rated on a Likert scale from 1 (disagree a lot) to 5 (agree a lot). An example of an item from this scale is as follows: "When I am active I find it pleasurable." In this study, the PACES had good internal consistency (Cronbach's α = 0.88).
Two scales were used to gauge affect in conjunction with the exercise tasks: the FS and the AD ACL. These scales were selected because they have been used in similar studies of the affective response to exercise and are sensitive to the changes in affect that accompany exercise (14,24,29). The FS (19) is a single-item 11-point bipolar measure of pleasure-displeasure. The scale ranges from −5 (very bad) to +5 (very good). A modified AD ACL also was used to gauge affect before and after each exercise task (31,32). The AD ACL is a 20-item self-report measure that assesses two arousal dimensions: energetic arousal (EA) and tense arousal (TA). Response options range from 1 (definitely feel) to 4 (definitely do not feel). The reliability and construct validity of the AD ACL are well established (31). To update the instrument for our contemporary adolescent sample, we replaced two adjectives on the TA scale (i.e., we replaced "clutched-up" with "nervous" and "intense" with "stressed"). Items within each subscale were averaged to provide a summary score for each subscale. In this study, the EA and TA subscales of the AD ACL had good internal consistency (Cronbach's α = 0.72 and α = 0.76).
Borg's (5) RPE scale was used as a manipulation check during the exercise tasks. This scale ranges from 6 (no exertion at all) to 20 (maximal exertion). Participants were asked to provide a rating on the RPE scale every three min throughout the exercise tasks.
To demonstrate that the two exercise tasks successfully produced significantly different levels of exertion, repeated-measures ANOVA examined the impact of the exercise on the periodic assessments of heart rate and RPE. Zero-order correlations were used to assess the relationship of the BIS and BAS scales to variables assessing exercise enjoyment, preference, and tolerance. The P values presented for zero-order correlations with BIS and BAS have been corrected for multiple comparisons using the Bonferonni method. Only those BAS scales that were significantly correlated with enjoyment of exercise were included in the subsequent repeated-measures analyses.
The relationships of BIS and BAS to the affective response to exercise were assessed using three separate intensity by time repeated-measures analyses. In each analysis, a measure of affect (FS, EA, and TA) was examined across all of the available time points after controlling for cardiovascular fitness. The BIS and BAS variables were transformed into dichotomous variables representing high and low scorers, and these dichotomous variables were entered into the repeated-measures analyses as between-subjects factors. For the BIS, a median split yielded two groups of low- (n = 73) and high- (n = 68) scoring participants, respectively. For the BAS(reward), the median split yielded two groups of low- (n = 63) and high- (n = 83) scoring participants, respectively. For FS, there were five time points (baseline, min 10, 20, 30, 40), and for the AD ACL scales there were three time points (baseline, min 30, 40). Initially, gender was included as a covariate in all analyses, but because there were no main or interactive effects for gender, it was dropped from the analyses.
Table 1 shows the means and SD for BIS, BAS, V˙O2peak, percent body fat, enjoyment, preference, and tolerance. Boys and girls were significantly different on several of the variables; boys had lower BIS and percent body fat, and higher preference, tolerance, and V˙O2peak.
Table 2 provides the correlations between the variables related to physical activity and fitness. Higher V˙O2peak was associated with lower percent body fat and with higher preference and tolerance. The measures of enjoyment, preference, and tolerance were all positively correlated with one another.
Exercise Task Work Rate
During the moderate-intensity exercise task, 96% of the study participants were able to maintain the target work rate throughout the 30-min exercise task. Five participants exhibited signs of fatigue during the task, resulting in an average reduction in work rate of 28 W (SD = 17) per participant for these five individuals (range = 10-55 W). During the hard-intensity exercise task, 85% of study participants exhibited signs of fatigue before the end of the 30-min task, resulting in a mean reduction in the target work rate of 33 W (SD = 16) per participant for these individuals (range = 5-80 W).
A 2 (exercise intensity) by 11 (time points: baseline, min 3, 6, 9, 12, 15, 18, 21, 24, 27, 30) repeated-measures ANOVA on the heart rate data showed a significant main effect of exercise intensity [F(1, 106) = 536.37, P < 0.001], time [F(10, 97) = 320.37, P < 0.001], and the intensity by time interaction [F(10,97) = 47.67, P < 0.001]. Figure 1 illustrates the pattern. During the moderate-intensity task, average heart rate increased from 76.40 bpm (SD = 13.99) during the warm-up to 137.00 bpm (SD = 18.98) at 18 min and then remained fairly stable throughout the rest of the 30 min. Average heart rate during the hard-intensity exercise task increased from 76.60 bpm (SD = 15.65) during the warm-up to 179.00 bpm (SD = 14.04) at 9 min and then remained stable for the rest of the task.
A 2 (exercise intensity) by 10 (time points: min 3, 6, 9, 12, 15, 18, 21, 24, 27, 30) repeated-measures ANOVA on the RPE data showed a significant main effect of exercise intensity [F(1,125) = 449.47, P < 0.00], time [F(9,117) = 76.74, P < 0.001], and the intensity by time interaction [F(9,117) = 11.63, P < 0.001]. Figure 2 shows that the average RPE during the moderate-intensity exercise task increased from 8.90 (very light; SD = 2.05) at 3 min to 13.60 (somewhat hard; SD = 3.28) at 30 min, with most of the increase occurring during the first 15 min of the task. Similarly, during the hard-intensity exercise task, the average RPE increased from 12.20 (somewhat hard; SD = 2.13) at 3 min to 17.00 (very hard; SD = 2.07) at 15 min and then remained fairly constant throughout the rest of the task.
Associations of BIS/BAS with Physical Fitness and Related Constructs
There was a significant negative correlation between the BIS and V˙O2peak (r = −0.26, P < 0.05) and between the BIS and tolerance (r = −0.27, P < 0.05). Adolescents who scored higher on the BIS scale had lower V˙O2peak and tolerance. The BAS(reward) subscale was positively related to enjoyment (r = 0.44, P < 0.05). None of the other BAS subscales was significantly associated with any of the fitness or exercise-related variables. The BIS scale was unrelated to any of the BAS scales.
Associations of BIS/BAS with Affective Response to Exercise
Because only the BAS(reward) subscale had a significant correlation with enjoyment, this scale was used in the repeated-measures analysis. The 2 (exercise intensity) by 5 (time points: baseline, min 10, 20, 30, 40) by 2 (BAS(reward): high and low) by 2 (BIS: high and low) repeated-measures ANOVA on the ratings of FS showed a main effect of intensity [F(1,137) = 19.76, P < 0.001], a main effect of time [F(4,134) = 6.63, P < 0.001], an interaction between time and intensity [F(4,133) = 3.00, P < 0.05], a main effect of the BIS (F = 11.05, P < 0.01), and a main effect of the reward subscale of the BAS (F = 9.02, P < 0.01). Post hoc t-tests revealed that, compared to adolescents who scored below the median on the BIS, adolescents who scored above the median on the BIS reported a lower FS at baseline (t = 2.10, P < 0.05), 10 min (t = 2.29, P < 0.05), 20 min (t = 2.11, P < 0.05), 30 min (t = 2.20, P < 0.05), and 40 min (marginally; t = 1.95, P =.05), regardless of the exercise intensity. Figure 3 shows the consistently lower BIS ratings across time at both exercise intensities; all t-tests comparing FS across levels of the BIS within each intensity were statistically significant. Similarly, compared to adolescents who scored below the median on the BAS(reward), adolescents who scored above the median on theBAS(reward) reported a higher FS at baseline (t = 2.21, P < 0.05), 10 min (t = 2.11, P < 0.05), 30 min (t = 2.07, P < 0.05), and 40 min (t = 3.44, P < 0.01), regardless of the exercise intensity. Figure 4 shows that the difference between FS ratings seems to be attenuated during the hard-intensity task; all t-tests comparing FS across levels of theBAS(reward) within each intensity were statistically significant at the P < 0.05 level except those for the hard-intensity task at baseline and 20 min (t-tests at 10 and 30 min during the hard-intensity task were marginally significant, P values < 0.10).
The 2 (exercise intensity) by 3 (time points: baseline, min 30, 40) by 2 (BAS(reward): high and low) by 2 (BIS: high and low) repeated-measures ANOVA on the ratings of EA showed a main effect of intensity [F(1,136) = 13.41, P < 0.001], a significant interaction between V˙O2peak and intensity [F(1, 136) = 4.64, P < 0.05], a main effect of the reward subscale of the BAS (F = 13.51, P < 0.001), and a three-way interaction between BAS(reward), intensity, and time [F(2, 135) = 3.19, P < 0.05]. Post hoc t-tests indicated that high BAS(reward) scorers reported feeling significantly more energy than low BAS(reward) scorers at the baseline (t = 2.50, P < 0.01), 30 min (t = 2.76, P < 0.01), and 40 min (t = 3.83, P < 0.001) time points for the moderate-intensity exercise task. In contrast, high BAS(reward) scorers reported significantly more energy than low BAS(reward) scorers at baseline (t = 3.27, P < 0.01) but not at 30 min (t = 1.58, P = 0.11), or 40 min (t = 1.72, P = 0.08) for the hard-intensity exercise task.
The 2 (exercise intensity) by 3 (time points: baseline, min 30, 40) by 2 (BAS(reward): high and low) by 2 (BIS: high and low) repeated-measures ANOVA on the ratings of TA showed no main or interactive effects of time, intensity, BIS, or BAS(reward) on ratings of TA. Similarly, the 2 (exercise intensity) by 10 (time points: baseline, min 3, 6, 9, 12, 15, 18, 21, 24, 27, 30) by 2 (BAS(reward): high and low) by 2 (BIS: high and low) repeated-measures ANOVA on RPE showed no main or interactive effects of time, intensity, BIS, or BAS(reward) on ratings of RPE.
The results of this study provide support for the hypotheses, though in a more nuanced fashion than anticipated. The hypothesis that both the BIS and the BAS would be correlated with physiological indicators of physical activity as well as exercise enjoyment and preference and tolerance for high-intensity exercise was partially confirmed. Only the BIS was correlated (negatively) with V˙O2peak and tolerance for high-intensity exercise, whereas only the BAS(reward) was correlated (positively) with exercise enjoyment. The hypothesis that the BIS would be associated with a more negative affective response to exercise, regardless of intensity, was supported by the results of the analyses using FS but not by the analyses using the AD ACL scales. Finally, the hypothesis that BAS would be related to a more positive affective response to exercise during the moderate task only was confirmed. High BAS(reward) scorers did report more positive affect in response to the exercise tasks, and the association was attenuated during the hard-intensity task, especially using the EA scale as the measure of affect.
The items on the BIS scale are designed to assess sensitivity to cues of punishment, and the results of this study suggest that adolescents with greater sensitivity to cues of impending punishment are likely to take less pleasure in exercising, to tolerate high-intensity exercise less well, and to have lower cardiovascular fitness as compared to adolescents with a lower sensitivity to cues of punishment. This finding suggests that high-BIS adolescents are at an elevated risk for becoming and remaining sedentary and may require special attention if they are to maintain adequate levels of physical activity for optimal fitness. The fact that the difference in affective responding to an exercise bout persisted even during the moderate-intensity exercise session is of particular concern because it suggests that high-BIS adolescents will find physical activity aversive even at the lower intensities usually assumed to be easily tolerated.
The items on the BAS(reward) scale are designed to assess positive responses to the anticipation of reward, and the results of this study indicate that adolescents with a greater sensitivity to cues of impending reward are more likely to report enjoying exercise and are also more likely to experience pleasure during exercise. It seems that, at a hard level of exercise intensity, this predisposition to experience positive affect with exercise may be obscured by the negative physiological cues that tend to accompany hard-intensity exercise (e.g., shortness of breath, perspiration, muscle strain, increased heart rate, etc.). Nevertheless, the BAS(reward) does seem to have a pervasive positive effect on the subjective experience of exercise because adolescents with high BAS(reward) scores reported more positive affect even in anticipation of the exercise task (at baseline) and after the cessation of the hard-intensity exercise task. Moreover, higher BAS(reward) scores were positively correlated with self-reported enjoyment of exercise. These findings suggest that high-BAS(reward) adolescents may be well-positioned to maintain participation in physical activity.
To our knowledge, this is the first study to examine the contribution of BIS/BAS to individual differences in affective response to exercise. The results suggest that both BAS and BIS have independent effects on the subjective experience of exercising. They also demonstrate that the effect of exercise is not uniform across individuals; deviations in personality characteristics may increase or decrease the likelihood that a person will experience a positive affective response to exercise. Thus, many of the studies that have examined mean differences in affective response to physical activity may have been hampered by a failure to take into account individual differences in affective responses.
Unlike the study by Hall et al. (18), we did not find a relationship between BIS and BAS on the one hand and RPE on the other. The explanation for this divergence may reside in the nature of the samples used in the two studies. We studied adolescents with an average age of 14 yr, whereas Hall et al. recruited college students with an average age of 21 yr. Adolescents at the younger age may still be undergoing physical changes associated with the pubertal growth spurt and may therefore be less accustomed to the physiological sensations that accompany a given level of exertion. In contrast, young adults may have a greater reservoir of experience in interpreting the sensations that accompany exercise and may therefore be more reliable in their perceptions of exertion. In other words, one might expect greater random variability in responses among adolescents, thus introducing statistical "noise" into the data and making any correlation more difficult to discern. Another difference between the two samples is that the adolescents in our study were noticeably less fit than the adults in the Hall et al. (18) study, as determined by reported levels of cardiovascular fitness. It may be that one would be more likely to find a relationship between BIS/BAS and RPE among individuals who are relatively fit. It is also worth noting that the absence of an association between BIS and RPE in our study is yet another argument that perceptions of exertion and affect are distinct constructs and, if one wants to comprehend adolescents' exercise experiences, the inclusion of affect seems necessary.
The finding that BIS was related to cardiovascular fitness, whereas BAS was not, is worthy of some discussion. It is relevant to note that BIS and BAS are often described as independent systems (7), with all combinations of low and high BIS/BAS being possible. Thus, an individual who scores high on the BIS scale may score either high or low on the BAS scale; they are not inversely related. Our findings suggest that the strength of the BIS has greater implications for fitness-related behavior. That is, although high-BAS(reward) adolescents reported greater enjoyment of exercise, that enjoyment did not seem to translate into higher fitness levels. These findings suggest that a promising strategy for promoting physical activity among adolescents would be to develop targeted, innovative strategies designed to encourage high-BIS adolescents to learn to enjoy exercise, perhaps by pairing physical activity with other preferred activities, such as active video games or social interaction. Future research should strive to test novel interventions that may be successful in motivating high-BIS adolescents to maintain physical activity participation.
The inclusion of two measures of affect in this study provides an opportunity to parse this rather omnibus variable into its components. According to the circumplex model of affect (15), there are two dimensions along which affect can vary: valence (pleasure-displeasure) and activation (low-high). The FS item, which asks the respondent to rate feelings on a single dimension from "bad" to "good," is an example of a tool that assesses affective valence without regard to activation. The two scales of the AD ACL used in this study are designed to tap into the two quadrants of the circumplex model of affect that correspond to activated pleasure (EA) and activated displeasure (TA). These are the two quadrants that are typically assessed concerning the affective response to exercise because physical activity is expected to increase activation (14,31).
Using the FS item, we found that BIS was related to the affective response to exercise, yet this association was absent with the subscales of the AD ACL (EA and TA). Our results suggest that the BIS is related to the pleasure-displeasure dimension of affect during an exercise bout, but not to the compound measures of affect that also incorporate the activation dimension. We speculate that, because the pleasure-displeasure dimension of affect is orthogonal to the activation-deactivation dimension, combining the two into a single assessment may introduce nonsystematic variability and therefore obscure the association between the BIS and affective valence.
Unlike the pattern of findings with respect to the BIS, the association between BAS(reward) and affective response to exercise emerged when either the FS or the EA was used to gauge affect, and did not emerge with respect to TA. To understand this discrepancy, it is helpful to identify the specific adjectives that are included in the EA and TA scales. EA is assessed with the following adjectives: active, energetic, vigorous, lively, and full-of-pep. TA is assessed with the following adjectives: jittery, stressed, fearful, anxious, and tense. As noted earlier, the BAS(reward) reflects sensitivity to cues of reward, so it makes intuitive sense that it would be related both to a measure of affective valence and to a measure of activated positive affect (energy) but not to activated negative affect (tension). In other words, the positivistic bias that may be introduced into adolescents' perception when the BAS(reward) is highly activated might be expected to yield differences in a measure of activation that is on the "pleasure" end of the valence continuum but not in a measure of activation that is on the "displeasure" end of the valence continuum.1
This study is limited by its cross-sectional design. The associations that emerged in this analysis are correlations, and the direction of the associations cannot be determined from the present data. In other words, it could be that individuals who experience a more positive emotional response to exercise also engage in greater amounts of physical activity and that this increased activity somehow stimulates the personality traits that are assessed by the BIS/BAS instrument. To disentangle these variables, it would be necessary to conduct a longitudinal study, in which BIS/BAS scores at one point in time are used to predict behavior change at a later point in time. Another limitation of this study is the reliance on self-report measures that have not been validated for use with adolescents. The FS, AD ACL, and the BIS/BAS scales have not undergone rigorous validation analysis with adolescent samples, although a recent study with adolescents (29) used the FS during acute exercise and reported similar patterns of responding as were found in the present study. All of these instruments have, however, been validated among adults, have excellent face validity, and demonstrated good internal consistency in this study. Still, it must be noted that we cannot assume that they are indeed measuring what we have set forth to measure. Finally, it should be noted that the effect sizes of the associations were generally small, in the range of 5% of the variance in affect explained. Clearly, there are other factors contributing to individual variability in affect during exercise. Future research should attempt to identify these additional factors, which may include cognitive variables (e.g., beliefs about the emotional benefits of exercise), social variables (e.g., concerns about social presentation in an exercise setting), and physiological variables (e.g., individual differences in neurophysiology).
The authors thank Lisa Bouchard for her contributions to completing the study and two anonymous reviewers for their very helpful comments. This work was supported in part by National Institutes of Health grant HD-37746 and the UCI Satellite GCRC (MO1 RR00827). The results of the present study do not constitute endorsement by ACSM.
1. Alfano CM, Klesges RC, Murray DM, Beech BM, McClanahan BS. History of sport participation in relation to obesity and related health behaviors in women. Prev Med
2. Azevedo MR, Araujo CL, Cozzensa da Silva M, Hallal PC. Tracking of physical activity
from adolescence to adulthood: a population-based study. Rev Saude Publica
3. Beck A, Steer R, Brown G. Manual for Beck Depression Inventory-II
. San Antonio (TX): Psychological Corporation; 1996. p. 38.
4. Blair C, Peters R, Granger D. Physiological and neuropsychological correlates of approach/withdrawal tendencies in preschool: further examination of the behavioral inhibition
system scales for young children. Dev Psychobiol
5. Borg G. Ratings of perceived exertion and heart rates during short-term cycle exercise and their use in a new cycling strength test. Int J Sports Med
6. Brenner SL, Beauchaine TP, Sylvers PD. A comparison of psychophysiological and self-report measures of BAS and BIS activation. Psychophysiology
7. Carver CS, Sutton S, Scheier M. Action, emotion, and personality: emerging conceptual integration. Pers Soc Psychol Bull
8. Carver CS, White TL. Behavioral inhibition
, behavioral activation
, and affective responses to impending reward and punishment: the BIS/BAS scales. J Pers Soc Psychol
9. Caspersen CJ, Pereira MA, Curran KM. Changes in physical activity
patterns in the United States, by sex and cross-sectional age. Med Sci Sports Exerc
10. Chapman L, Chapman J. The measurement of handedness. Brain Cogn
11. Cooper A, Gomez R, Aucote H. The Behavioral Inhibition
System and Behavioral Approach System (BIS/BAS) Scales: measurement and structural invariance across adults and adolescents. Pers Individ Differ
12. Eaton DK, Kann L, Ross J, Hawkins J, Harris WA, et al. Youth risk behavior surveillance-United States, 2005. MMWR Surveill Summ
13. Ekkekakis P, Hall EE, Petruzzello SJ. Practical markers of the transition from aerobic to anaerobic metabolism during exercise: rationale and a case for affect
-based exercise prescription. Prev Med
14. Ekkekakis P, Hall EE, Petruzzello SJ. The relationship between exercise intensity and affective responses demystified: to crack the 40-year-old nut, replace the 40-year-old nutcracker! Ann Behav Med
15. Ekkekakis P, Petruzzello SJ. Analysis of the affect
measurement conundrum in exercise psychology; a conceptual case for the affect
circumplex. Psychol Sport Exerc
16. Ekkekakis P, Thome J, Petruzzello SJ, Hall EE. The Preference for and Tolerance of the Intensity of Exercise Questionnaire: a psychometric evaluation among college women. J Sports Sci
17. Franken IH, Muris P, Georgieva I. Gray's model of personality and addiction. Addict Behav
18. Hall EE, Ekkekakis P, Petruzzello SJ. Is the relationship of RPE to psychological factors intensity-dependent? Med Sci Sports Exerc
19. Hardy C, Rejeski W. Not what, but how one feels: the measurement of affect
during exercise. J Sport Exerc Psych
20. Hills A, King N, Armstrong T. The contribution of physical activity
and sedentary behaviors to the growth and development of children and adolescents: implications for overweight and obesity. Sports Med
21. Hu G, Lakka T, Kilpelainen T, Tuomilehto J. Epidemiological studies of exercise in diabetes prevention. Appl Physiol Nutr Metab
22. Kasch KL, Rottenberg J, Arnow BA, Gotlib IH. Behavioral activation
and inhibition systems and the severity and course of depression. J Abnorm Psychol
23. Kendzierski D, DeCarlo K. Physical Activity
Enjoyment Scale: two validation studies. J Sport Exerc Psychol
24. Lochbaum MR. Viability of resting electroencephalograph asymmetry as a predictor of exercise-induced affect
: a lack of consistent support. J Sport Behav
25. Motl RW, Dishman RK, Saunders R, Dowda M, Felton G, Pate RR. Measuring enjoyment of physical activity
in adolescent girls. Am J Prev Med
26. National Center for Health Statistics. Health, United States, 2004. In: Centers for Disease Control and Prevention, editor. Vol. DHHS Publication Number 2004-1232, Series Health, United States, 2004
. US Department of Health and Human Services; 2004. p. 255.
27. Rhodes R. The built-in environment: the role of personality and physical activity
. Exerc Sport Sci Rev
28. Rhodes R, Smith N. Personality correlates of physical activity
: a review and met-analysis. Br J Sports Med
29. Sheppard K, Parfitt G. Acute affective responses to prescribed and self-selected exercise intensities in young adolescent boys and girls. Pediatr Exerc Sci
30. Swinburn B, Caterson I, Seidell J, James WP. Diet, nutrition and the prevention of excess weight gain and obesity. Public Health Nutr
31. Thayer RE. Activation-deactivation checklist: current overview and structural analysis. Psychol Rep
32. Thayer RE. The Biopsychology of Mood and Arousal
. New York (NY): Oxford University Press; 1989.
33. Troiano R, Berrigan D, Dodd K, Mâsse LC, Tilert T, McDowell M. Physical activity
in the United States measured by accelerometer. Med Sci Sports Exerc
34. Wannamethee S, Shaper A. Physical activity
and cardiovascular disease. Semin Vasc Med
35. Whipp B, Davis J, Torres F, Wasserman K. A test to determine parameters of aerobic function during exercise. J Appl Physiol