Exercise training during adjuvant therapy for breast cancer can improve outcomes (9,18,20,21), but adherence is difficult because of the stresses of a cancer diagnosis and the well-known side effects of treatments (6). Previous studies of breast cancer patients receiving adjuvant therapies have reported exercise adherence rates below 80% (9,18,20,21) but have not determined the predictors of such adherence (6). Identifying the key predictors of exercise adherence may help in the development of targeted behavioral interventions that achieve better outcomes in this clinical setting.
We conducted the Supervised Trial of Aerobic versus Resistance Training (START) comparing aerobic exercise training (AET) and resistance exercise training (RET) to usual care (UC) in breast cancer patients initiating adjuvant chemotherapy (7-9). We previously reported that exercise training during chemotherapy improved self-esteem, aerobic fitness, muscular strength, lean body mass, body fat levels, and chemotherapy completion rate at postintervention (9) and also improved self-esteem and anxiety at 6 months follow-up (7). We achieved an exercise adherence rate of 70.2% during the trial and observed positive associations between exercise adherence and several improved outcomes including aerobic fitness, muscular strength, lean body mass, and chemotherapy completion rate (9). These data suggest that our exercise training interventions were effective but that better adherence may lead to even better outcomes. The purpose of the present study was to identify the key predictors of supervised exercise training during the trial.
Our investigation was guided by Ajzen's (1) theory of planned behavior (TPB). The TPB is a social cognitive model of human behavior that proposes that intention (i.e., motivation) is the most important determinant of behavior. Intention, in turn, is influenced by perceived behavioral control (i.e., perceived ease or difficulty of performing the behavior), attitude (i.e., a positive or negative evaluation of performing the behavior), and subjective norm (i.e., perceived approval/support for performing the behavior). The TPB has demonstrated utility in predicting exercise behavior in the general population (11) and in cancer survivors (6). We also examined standard demographic (e.g., age, marital status), medical (e.g., disease stage, treatments), behavioral/fitness (e.g., recent exercise, smoking, aerobic fitness, muscular strength), and patient-rated variables (e.g., quality of life, depression) (25). The TPB proposes that such "external" variables influence human behavior through their effects on social cognitive beliefs. Consequently, we hypothesized that the TPB variables would be the strongest predictors of exercise adherence during the trial and that any associations between the "external" variables and exercise adherence would be mediated by such beliefs.
Setting and participants.
The methods of the START trial have been reported elsewhere (7-9). Here, we briefly summarize the methods with additional information on the TPB variables. Participants were being treated at the Cross Cancer Institute in Edmonton, Alberta, the Ottawa Hospital Integrated Cancer Program in Ottawa, Ontario, and the British Columbia Cancer Agency in Vancouver, British Columbia, Canada. Ethical approval was obtained from each center, and written informed consent was obtained from each participant. Eligibility criteria included women ≥18 yr old with stage I-IIIA breast cancer initiating adjuvant chemotherapy. Women were excluded if they had incomplete axillary surgery, transabdominal rectus abdominus muscle reconstructive surgery, uncontrolled illnesses, completed more than one cycle of chemotherapy, or were not approved by their oncologist. Women were not excluded based on past or recent exercise levels.
Design and procedures.
The study was a prospective, three-armed, randomized controlled trial. Eligible participants were identified by their treating oncologist and approached by either the oncologist or the research coordinator. If interested, participants received a questionnaire package and a follow-up telephone call to confirm eligibility and interest. Participants who agreed to participate in the trial were then booked for exercise testing, a body composition assessment, and a blood draw. After completing all baseline testing, participants were stratified by cancer center and chemotherapy regimen (taxane-based vs non-taxane-based) and randomly assigned to AET, RET, or UC using a computer-generated program with allocation sequence concealed from the project directors who assigned participants to groups.
Exercise training interventions.
Participants assigned to AET or RET were asked to exercise for the duration of their chemotherapy (defined as beginning within 1-2 wk after their first chemotherapy cycle and ending 3 wk after their final chemotherapy cycle). All exercise sessions were supervised at well-equipped fitness centers by qualified staff. All three fitness centers were open 8:00 a.m. to 5:00 p.m. Monday to Friday with Edmonton and Vancouver making additional accommodations for evenings and weekends if needed. The AET group was asked to exercise three times per week on a cycle ergometer, treadmill, or elliptical trainer beginning at 60% of their V˙O2peak for 15 min and progressing to 80% of their V˙O2peak for 45 min. V˙O2peak was determined by a baseline symptom-limited maximal exercise test with gas exchange analysis. The RET group was asked to exercise three times per week performing two sets of 8-12 repetitions of nine different exercises at 60-70% of their estimated one repetition maximum (RM) based on results from a baseline 8 RM test. The weight was increased by 10% when participants could complete >12 repetitions. Exercise adherence was recorded by the fitness trainers including attendance, duration, and intensity. Given that more than 95% of the sessions attended were done at the appropriate duration and intensity, we calculated adherence as the number of sessions attended divided by the number of sessions expected based on the length of the chemotherapy protocol (which ranged from 12 to 24 wk).
Assessment of predictors.
Demographic data were collected by self-report and consisted of age, marital status (0 = not married; 1 = married), education (six categories ranging from 1 = some high school to 6 = completed graduate school), annual family income (six categories ranging from 1 = <$20,000 to 6 = >$100,000), employment status (0 = not employed full-time; 1 = employed full-time), and location/center (0 = Ottawa; 1 = Edmonton; 2 = Vancouver). Medical data were collected from medical records and consisted of disease stage (I, IIa, IIb, and IIIa coded as 0-3), type of surgery (0 = lumpectomy; 1 = mastectomy), and type of chemotherapy (nontaxane = 0; taxane = 1) based on higher rates of toxicities for taxanes including myalgias and peripheral neuropathies. Behavioral variables were collected from self-report and consisted of smoking (0 = nonsmoker; 1 = smoker) and exercise since the time of diagnosis (12) coded as meeting or not meeting the public health guidelines (14) of 150 min of moderate-to-vigorous exercise per week or 60 min of vigorous exercise per week.
Physical fitness variables included peak oxygen consumption (V˙O2peak) assessed using a maximal incremental exercise protocol on a treadmill with gas exchange analysis (26) and muscular strength assessed using an eight repetition maximum (RM) test on the leg extension to estimate 1RM (17). Body weight to the nearest 0.1 kg and standing height to the nearest 0.5 cm were assessed without shoes using a balance beam scale to calculate body mass index. A dual x-ray absorptiometry scan was obtained for the assessment of whole-body fat and lean tissue (added after the first 23 participants were randomized). Patient-rated outcomes (PRO) consisted of cancer-specific quality of life (QoL) and fatigue assessed by the Functional Assessment of Cancer Therapy-Anemia scale (3). We also assessed psychosocial functioning using the Rosenberg Self-Esteem Scale (23), the Center for Epidemiological Studies-Depression Scale (22), and the Spielberger State Anxiety Inventory (24).
TPB constructs were assessed before randomization by all participants for both aerobic and resistance exercise based on standard guidelines (2). Intention was measured by asking "How motivated are you to do the aerobic (resistance) exercise program?" with response options ranging from 1 (slightly motivated) through 4 (moderately motivated) to 7 (extremely motivated). Attitude was measured using two items that tapped instrumental (useful-useless, harmful-beneficial) and affective (unenjoyable-enjoyable, unpleasant-pleasant) attitudes, respectively. The stem for the attitude items was "I think that doing aerobic (resistance) exercise during my chemotherapy treatment will be…" with response options ranging from 1 (extremely "negative descriptor") to 7 (extremely "positive descriptor"). Perceived behavioral control was measured by two items. An example was: "If I wanted to, I could easily do the aerobic (resistance) exercise program during my chemotherapy" (1 = strongly disagree to 7 = strongly agree). Subjective norm was measured by a single item rated on a 7-point scale from 1 (strongly disagree) to 7 (strongly agree). The item was: "Most people who are important to me would support me doing aerobic (resistance) exercise during my chemotherapy treatment."
We compared the three randomized groups at baseline on the TPB variables using analyses of variance. We also compared participants beliefs about aerobic versus resistance exercise using dependent t-tests. We analyzed the univariate associations between the predictors and exercise adherence using Pearson correlations. For ease of interpretation, we also analyzed the continuous predictors as categorical variables based on clinically relevant or statistically determined cut-points using analyses of variance. Variables that had statistically significant or borderline significant (P < 0.10) univariate associations with exercise adherence were retained for a forced entry multivariate regression analyses.
Flow of participants through the trial has been reported elsewhere (9). Briefly, we screened 1468 breast cancer survivors and recruited 242 of 736 (33%) eligible participants between February 2003 and July 2005. We randomized 160 to the two exercise groups. Age ranged from 25 to 78 yr (mean, 49 yr), 21% were obese, 37% were postmenopausal, 61% had stage II disease, 59% received breast conservation surgery, 31% received a taxane-based chemotherapy, and 25% reported recent exercise. There were no differences among the groups. The median length of the exercise intervention was 17 wk (95% CI, 9-24), and adherence to the supervised exercise program during chemotherapy was 72.0% ± 30.1% and 68.2% ± 28.4% in the AET and RET groups, respectively (P = 0.411), with 51.2% achieving 80% attendance (range, 0% to 100%).
Baseline TPB variables are presented in Table 1. There were no differences among the randomized groups. Overall, our sample of breast cancer patients felt that exercise training during chemotherapy would be quite beneficial, slightly enjoyable, and moderately difficult. They also moderately to strongly agreed that most people would support them, and they were moderately to extremely motivated to exercise during their chemotherapy. In comparing beliefs about the two types of exercise training before randomization, breast cancer patients felt that aerobic exercise would be more beneficial than resistance exercise (P < 0.001), more enjoyable (P = 0.015), and they would receive more support for doing it (P = 0.001). They did not, however, believe it would be easier to do (P = 0.237) nor were they more motivated to do it (P = 0.517).
Predictors of adherence to the exercise training program.
Preliminary analyses indicated no interactions between the predictors and group assignment (AET vs RET) for predicting adherence. Consequently, both exercise groups were combined to increase the study power. Correlational and categorical analyses of the predictors of exercise adherence are presented in Table 2 (demographic and medical), Table 3 (behavioral and fitness), Table 4 (PRO), and Table 5 (TPB). We found statistically significant or borderline significant univariate associations between exercise adherence and location/center (r = 0.30; P < 0.001), V˙O2peak (r = 0.21; P = 0.008), muscular strength (r= 0.21; P = 0.008), percent body fat (r = −0.21; P = 0.012), disease stage (r = 0.17; P = 0.031), education (r = 0.15; P = 0.053), depression (r = −0.14; P = 0.073), and smoking (r = −0.14; P = 0.081). Higher exercise adherence was achieved by breast cancer patients that were in Vancouver, fitter, stronger, thinner, more advanced disease stage, better educated, less depressed, and not smoking.
The significant or borderline significant predictors obtained on the entire sample of 160 (excluding percent body fat because of the smaller sample size of 144) were forced into a multivariate regression analysis and explained 21% of the variance in exercise adherence. Independent predictors of exercise adherence in the model were location/center (β = 0.28; P = 0.001), V˙O2peak (β = 0.19; P = 0.016), disease stage (β = 0.18; P = 0.015), and depression (β = −0.16; P = 0.033). Muscular strength (β = 0.04; P = 0.666), education (β = 0.07; P = 0.360), and smoking (β = −0.08; P = 0.282) became nonsignificant. These results were largely unchanged after including percent body fat in the model despite the reduction in sample size to 144. With percent body fat in the model, the significant or borderline significant predictors explained 21% of the variance in exercise adherence with independent contributions being made by location/center (β = 0.24; P = 0.008), disease stage (β = 0.23; P = 0.004), depression (β = −0.19; P = 0.021), and V˙O2peak (β = 0.17; P = 0.119). Muscular strength (β = 0.03; P = 0.745), education (β = 0.03; P = 0.736), percent body fat (β = −0.07; P = 0.530), and smoking (β = −0.08; P = 0.328) became nonsignificant.
Adherence to supervised exercise training during chemotherapy was 70.2% in the START trial. This adherence rate is within the range reported for other trials of exercise during cancer treatments (18,20,21) and older adults without cancer (19). Importantly, however, many of these studies have reported adherence rates that excluded dropouts (i.e., participants who did not complete the trial). Martin and Sinden (19) reported that in exercise randomized controlled trials (RCTs) of older adults without cancer, the average adherence rate for studies that excluded dropouts was 88%, compared to 63% for studies that included dropouts. Given this comparison, our adherence rate is superior to that reported in RCTs of older adults, although it was still not optimal. Consequently, understanding the predictors of adherence in our trial is an important endeavor.
Contrary to our hypothesis, motivational variables did not predict exercise adherence. This finding is very rare in the exercise behavior literature in general (13) and in research with cancer patients and survivors (6). Few cancer studies, however, have examined the predictors of supervised exercise adherence during cancer treatments. In a previous trial of prostate cancer patients receiving androgen deprivation therapy (ADT), the independent predictors of supervised exercise adherence were exercise stage of change, intention, and age (10). In a previous trial of mixed cancer survivors on and off treatments attending group psychotherapy sessions, the independent predictors of self-reported exercise adherence were sex (being male), extraversion, normative beliefs (negative), and perceived behavioral control (5). Finally, in a previous trial of colorectal cancer survivors mostly on treatments, the independent predictors of self-reported exercise adherence were exercise stage of change, employment status (not working full time), treatment protocol (receiving less treatment), and perceived behavioral control (4).
There are several possible explanations for why motivational variables did not predict exercise adherence in the START trial. First, there was very limited variability in several of the TPB variables at baseline, likely attributable to the select and highly motivated sample of participants that volunteer for such a trial. It is well known that limited variability reduces possible associations among variables. Second, social cognitive beliefs generally only predict behavior when the beliefs are relatively accurate (2), which is usually the case when respondents have had previous experience with a behavior. Participants in our study had no previous experience with exercising during chemotherapy and so their beliefs at baseline were a "best guess" about what it might be like to exercise during chemotherapy. Third, a previous analysis of the most common exercise barriers in the START trial demonstrated that over half the barriers were related to disease- or treatment-related side effects and fewer than 15% appeared to reflect motivational issues (8). Consequently, exercise adherence during chemotherapy may be more impacted by uncontrollable factors such as treatment response and toxicities that accumulate over the course of chemotherapy than by motivation. Motivation, conversely, is likely to have a large influence on who might attempt such an exercise program in the first place (as evidenced by the positive TPB beliefs at baseline).
The strongest predictor of exercise adherence in our trial was the location/center at which the exercise was performed. Although this finding was unexpected, there are several possible explanations. First, the amount of personal attention received by participants may have varied across centers. In Vancouver, participants trained at a new fitness facility that, at the time, included only START trial participants. Consequently, each participant received one-on-one supervision with few other participants in the center at any one time. Edmonton exercisers trained at a similar facility and received similar attention early in the trial. About half way through the trial, however, another large trial was launched in Edmonton, which significantly increased the volume of exercisers in the center and reduced the amount of one-on-one attention. Conversely, Ottawa exercisers trained at a fitness facility located at the cancer center that was open to staff and patients not participating in clinical trials. This situation resulted in a much busier facility with less personal attention. Other contributing factors may have included (a) the more moderate climate in Vancouver compared to harsher winters in Edmonton and Ottawa, (b) more flexible hours of operation in Vancouver and Edmonton (including weekends, evenings, and holidays) compared to Ottawa (regular weekday hours only), (c) the SARS epidemic in Ottawa that resulted in the hospital and fitness center being closed for 2 wk, (d) possible demographic differences in each city, (e) different referral patterns at each cancer center, and (f) travel time and convenience of accessing each fitness center.
It is unclear why participants with more advanced disease stage, particularly stage IIIa, achieved better adherence in the trial. Several studies have reported inverse associations between disease stage and exercise behavior, although most have used cross-sectional surveys of posttreatment cancer survivors and relied on self-reported exercise behavior (6,16). In a previous trial of prostate cancer patients receiving ADT, higher disease stage predicted lower adherence to supervised exercise (10). In a trial of colorectal cancer survivors, tumor stage was negatively associated with self-reported exercise adherence in univariate analysis (4). It is possible that breast cancer patients with more advanced disease stage experienced fewer difficulties during chemotherapy; however, this would not be expected based on their chemotherapy protocols. Given the unexpected nature of this finding, replication is required before it can be considered reliable.
Fitter and stronger breast cancer patients also demonstrated better adherence, perhaps partly reflecting fewer underlying comorbidities. It is well known that past exercise is one of the best predictors of future exercise (25). It is intuitive that breast cancer patients who exercised before their chemotherapy would be more successful in adhering to exercise during chemotherapy, but we did not find this based on our self-report exercise measure. Recent exercise is likely unstable in newly diagnosed breast cancer patients and chemotherapy is a major intervening event that can alter exercise behavior. It is possible that objective measures of fitness and strength are better indicators of recent or past exercise than self-reported exercise or that a minimum fitness level is needed to allow breast cancer patients to tolerate and respond to exercise training during chemotherapy. Although the minimum fitness level needed is unknown, our data suggest that a V˙O2peak >20 mL·kg−1·min−1 may be a threshold. Below this fitness level is considered poor and well below the expected mean of 30 mL·kg−1·min−1 for women in the same age range in the general population (15). Future trials should determine if very unfit women can benefit from exercise during chemotherapy or whether it may be more prudent to initiate exercise training in these women after adjuvant therapies are completed. The fact that baseline aerobic fitness predicted exercise adherence is a dilemma for exercise researchers. Demonstrating the benefits of an exercise intervention is a precarious balance between enrolling participants who are most likely to benefit (i.e., generally those who are least fit) and enrolling participants who are most likely to adhere (i.e., generally those who are more fit).
Higher levels of depression also predicted poorer exercise adherence during chemotherapy, although no threshold could be identified. The association between poor psychological health and exercise behavior has been well established in various populations (25) and has recently been reported in breast cancer survivors (16). Clearly, depression at the start of cancer treatments is problematic for many reasons, and it is intuitive that it may also interfere with lifestyle changes.
It was reassuring to observe that age, type of surgery, chemotherapy protocol, QoL, fatigue, and anxiety did not predict adherence. These data suggest that older breast cancer patients and those receiving different surgical and chemotherapy interventions adhered equally well to exercise training during chemotherapy. Moreover, having poor QoL, being fatigued, or anxious at baseline did not interfere with exercise adherence during this difficult time. Previous studies have reported that age, QoL, and fatigue are associated with exercise participation during survivorship in cross-sectional surveys using self-reported exercise and they may predict adherence in clinical trials (6,16). The differences in findings could be attributable to many factors including the type of exercise (home-based/self-reported vs center-based/supervised), the type of cancer treatment, and/or the specific cancer survivor group.
The overall strengths of our trial have been noted elsewhere (7-9). Additional strengths specific to the present report include being the first study to prospectively examine the predictors of supervised exercise training in breast cancer patients receiving chemotherapy, the assessment of many different potential predictors of exercise adherence, the adoption of a validated theoretical model, and the use of an objective measure of exercise adherence. The overall limitations of our trial have also been noted elsewhere (7-9). A key limitation of the present report is that we focused on predictors of supervised exercise training in breast cancer patients receiving chemotherapy. Predictors of exercise adherence may differ based on type of cancer, type of treatments, location of exercise, and other factors. Moreover, we were only able to explain 21% of the variance in exercise adherence, leaving most of the variation unexplained, indicating that not all important predictors of exercise were assessed in the present study. Some likely prospects for missed important predictors include treatment toxicities and possible family and social roles provided by women.
In summary, we examined predictors of supervised exercise adherence in breast cancer patients receiving chemotherapy. Our exercise adherence rate was good, but not optimal, and it was predicted by a wide range of variables including geographic, demographic, medical, behavioral, fitness, and psychosocial. The independent predictors in a multivariate analysis were location/center, aerobic fitness, disease stage, and depression. Location/center in our trial may have been a proxy for the amount of one-on-one attention at the facilities, although we cannot rule out any number of other differences (e.g., presence of other clientele, weather, SARs, travel distance/convenience, referral patterns, etc.). Consideration should be given to these factors when designing behavioral support interventions in this clinical setting. Our results also suggest that motivational variables are not important predictors of adherence in breast cancer patients already motivated enough to volunteer for a supervised exercise trial. The present study provides an empirical basis for improving exercise adherence in breast cancer patients on chemotherapy.
The authors gratefully acknowledge Lisa Workman, MA; Neil Eves, PhD; John McGavock, PhD; Kristin Campbell, PhD; Margaret McNeely, BScPT, MSc; Diana Jespersen, RN; Chris Scott, BSc; Lianne Dolan, MSc; Ben Wilson, BSc; Christopher Sellar, MS; and Diane Cook, BPE, for their assistance with the trial. This study was funded by a grant from the Canadian Breast Cancer Research Alliance. KSC is supported by the Canada Research Chairs Program. RDR is supported by a New Investigator Award from the Heart and Stroke Foundation of Canada. CMF is supported by a Health Scholar Award from the Alberta Heritage Foundation for Medical Research (AHFMR). JKV was supported by a Canada Graduate Scholarship from Canadian Institutes of Health Research and an Incentive Award from AHFMR. The results of this study do not constitute endorsement by the American College of Sports Medicine.
1. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process
2. Ajzen I. Nature and operation of attitudes. Ann Rev Psychol
3. Cella D, Eton DT, Lai JS, Peterman AH, Merkel DE. Combining anchor and distribution-based methods to derive minimal clinically important differences on the Functional Assessment of Cancer Therapy (FACT) anemia and fatigue scales. J Pain Symptom Manage.
4. Courneya KS, Friedenreich CM, Quinney HA, Fields AL, Jones LW, Fairey AS. Predictors of adherence and contamination in a randomized trial of exercise in colorectal cancer survivors. Psycho-Oncol.
5. Courneya KS, Friedenreich CM, Sela RA, Quinney HA, Rhodes RE. Correlates of adherence and contamination in a randomized controlled trial of exercise in cancer survivors: an application of the theory of planned behavior and the five factor model of personality. Ann Behav Med.
6. Courneya KS, Karvinen KH, Vallance JKH. Exercise motivation and behavior change. In: Handbook of Cancer Survivorship
, M Feuerstein. New York (NY): Springer, 2007, pp. 113-32.
7. Courneya KS, Segal RJ, Gelmon K, et al. Six-month follow-up of patient-rated outcomes in a randomized controlled trial of exercise training during breast cancer chemotherapy. Cancer Epidemiol Biomarkers Prev.
8. Courneya KS, McKenzie DC, Reid RD, Mackey JR, et al. Barriers to supervised exercise training in a randomized controlled trial of breast cancer patients receiving chemotherapy. Ann. Behav. Med
9. Courneya KS, Segal RJ, Mackey JR, et al. Effects of aerobic and resistance exercise in breast cancer patients receiving adjuvant chemotherapy: a multicenter randomized controlled trial. J Clin Oncol.
10. Courneya KS, Segal RJ, Reid RD, et al. Three independent factors predicted adherence in a randomized controlled trial of resistance exercise training among prostate cancer survivors. J Clin Epidemiol.
11. Downs D, Hausenblas HA. Exercise behavior and the theories of reasoned action and planned behavior: a meta-analytic update. J Phys Activ Health
12. Godin G, Shephard RJ. A simple method to assess exercise behavior in the community. Can J Appl Sport Sci
13. Hagger M, Chatzisarantis N, Biddle S. A meta-analytic review of the theories of reasoned action and planned behavior in physical activity: predictive validity and the contribution of additional variables. J Sport Exercise Psychol.
14. Haskell WL, Lee I, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc.
15. Heyward V. Assessing cardiorespiratory fitness. In: Advanced Fitness Assessment & Exercise Prescription
, Wikfren S, Mustain E, Merz Bott S, Cler E, Graham M. Champaign (IL): Human Kinetics, 1998, pp. 47-81.
16. Hong S, Bardwell WA, Natarajan L, et al. Correlates of physical activity level in breast cancer survivors participating in the Womens Healthy Eating and Living (WHEL) Study. Breast Cancer Res.
17. Lander J. Maximum based on reps. Natl Strength Cond Assoc J.
18. Markes M., T. Brockow, and K. L. Resch. Exercise for women receiving adjuvant therapy for breast cancer. Cochrane Database of Systematic Reviews. 2006;4.
19. Martin K, Sinden A. Who will stay and who will go? A review of older adults' adherence to randomized controlled trials of exercise. J Aging Phys Activ.
20. McNeely ML, Campbell KL, Rowe BH, Klassen TP, Mackey JR, Courneya KS. Effects of exercise on breast cancer patients and survivors: a systematic review and meta-analysis. CMAJ
21. Mutrie N, Campbell A, Whyte F, et al. Benefits of supervised group exercise program for women being treated for early stage breast cancer: pragmatic randomised controlled trial. BMJ
22. Radloff L. The use of the center for epidemiologic studies depression scale in adolescents and young adults. J. Adolesc.
23. Rosenberg M. Society and the Adolescent Self-image
, Princeton (NJ): Princeton University Press, 1965, p. 326.
24. Spielberger C, Gorusch R, Lushene R. Manual for the State-Trait Anxiety Inventory
, Palo Alto (CA): Consulting Psychologists Press, 1970, p. 257.
25. Trost SG, Owen N, Bauman AE, Sallis JF, Brown W. Correlates of adults' participation in physical activity: review and update. Med Sci Sports Exerc.
26. Wasserman K, Hansen J, Sue D, Casaburi R, Whipp B. Principles of Exercise Testing and Interpretation: Including Pathophysiology and Clinical Applications
, 3rd ed, Philadelphia (PA): Lippincott, Williams & Wilkins, 1999, p. 556.
Keywords:©2008The American College of Sports Medicine
CANCER SURVIVORS; DETERMINANTS; PHYSICAL ACTIVITY; RANDOMIZED CONTROLLED TRIAL