UvA-DARE (Digital Academic Repository) Physical exercise during adjuvant chemotherapy van Waart, H. Link to publication

UvA-DARE (Digital Academic Repository) Physical exercise during adjuvant chemotherapy van Waart, H. Link to publication Citation for published version (APA): van Waart, H. (2017). Physical exercise during adjuvant chemotherapy General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl) Download date: 06 Dec 2017

Why do patients choose (not) to participate in an exercise trial during adjuvant chemotherapy for breast cancer? Hanna van Waart Wim H. van Harten Laurien M. Buffart Gabe S. Sonke Martijn M. Stuiver Neil K. Aaronson Psycho-Oncology, 2016; 25: 964-970 dx.doi.org/10.1002/pon.3936

ABSTRACT Objective Only between 25% and 50% of patients invited to participate in clinical trial-based physical exercise programs during cancer treatment agree to do so. The purpose of this study was to identify factors associated significantly with the decision (not) to participate in a randomized controlled trial of physical exercise during adjuvant chemotherapy for breast cancer. Methods Based on questionnaire data, we compared trial participants and non-participants on a range of sociodemographic, clinical health-related, practical, behavioral, and attitudinal variables. Results Two hundred thirty of 524 patients agreed to participate in the trial (44%). The 294 (56%) non-participants indicated that they wanted to exercise on their own or that they did not wish to exercise in the context of a trial. Those who preferred to exercise on their own were relatively similar to trial participants but were more likely to be in the maintenance exercise stage. Those non-participants who did not wish to exercise had a significantly lower level of education, were less likely to be working, reported more fatigue and lower health-related quality of life, had lower sense of self-efficacy, more negative attitudes towards exercise, less social support, and perceived fewer benefits and more barriers to exercising during treatment than trial participants. Conclusion Minimizing practical barriers to participation, providing educational materials on the potential benefits of exercise, and giving adequate professional and social network encouragement may increase the number of patients willing to exercise during treatment and to participate in such studies. 96

Non-participation INTRODUCTION A rapidly increasing number of studies have assessed the effects of physical exercise in patients with cancer. Systematic reviews support that exercise reduces fatigue, 1 and improves physical fitness 2 and health-related quality of life. 3 In many cases, only a small percentage of eligible patients participate in exercise trials. It is likely that those patients who participate are not entirely representative of the larger target population of interest. The observed effects in exercise trials may overestimate or underestimate the true effect in the target population, depending on the reasons for non-participation. There is a range of reasons why patients may not wish to participate in exercise trials. Previous studies among varied patient groups have identified logistic barriers (e.g., difficulty fitting exercise into one s daily schedule), health-related barriers (e.g., side effects of treatment), and attitudinal barriers (e.g., aversion to or low perceived benefit of exercising) or trial-specific reasons (e.g., not wanting to risk assignment to a control group). 4 8 To the best of our knowledge, no studies, to date, have investigated reasons for non-participation in exercise-oncology trials. Increased understanding of the characteristics of (non-)participants in exercise trials may help us understand the nature of bias that may be introduced due to (selective) nonparticipation. It can also provide clues as to how we might tailor interventions to reach those patients who are reluctant or unwilling to participate in exercise interventions. In the Physical Exercise during Chemotherapy Effectiveness (PACES) trial, we demonstrated that exercising during chemotherapy has a significant, salutary effect on physical fitness, fatigue, and chemotherapy completion rates. 9 However, slightly more than half of the eligible patients declined to participate. The objective of the current, planned analysis of data from the PACES trial was to investigate the reasons for (non-)participation and to identify sociodemographic, clinical, health-related, practical, and attitudinal factors that are associated with (non-)participation. Several observational studies have shown that behavioral constructs, including the Theory of Planned Behavior, and the Transtheoretical Model, can be useful in explaining exercise behavior in cancer patients. 10,11 Based on these models, we hypothesized that trial participants would report better health, more social support, more positive attitudes towards physical exercise, higher self-efficacy, and fewer perceived barriers to and more benefits from being physically active than non-participants. 6 METHODS Study sample The PACES study was a randomized controlled trial that evaluated the efficacy of physical exercise in maintaining physical condition and minimizing fatigue during chemotherapy. 9 Participants were randomized to one of three groups: 1) Onco-Move, a low intensity, homebased physical activity program (intensity level of 12 14 on the Borg scale of perceived exertion 12 ); 2) OnTrack, a moderate-high intensity, combined supervised aerobic and 97

strength exercise program (aerobic intensity level of 50% to 80% of the maximal workload as estimated by the steep ramp test 13 ; strength intensity level of two times eight repetitions at 80% 1-repetition maximum); or 3) Usual Care control group. Patients were eligible if they had primary breast or colon cancer and were scheduled for adjuvant chemotherapy at one of 12 hospitals in the broader Amsterdam region of the Netherlands between April 2010 and December 2012. The current analysis is restricted to patients with breast cancer, who represented the large majority of those invited to participate in the trial (91%). Potentially eligible patients were screened for the presence of comorbid conditions that would contraindicate participation in a physical activity program. This excluded patients with serious orthopedic, cardiovascular, or cardiopulmonary conditions (or risks), those who suffered malnutrition, or had serious psychiatric or cognitive problems. For assessment purposes, study participants had to have basic fluency in Dutch. There was no upper age limit. Study procedures Eligible patients who declined to participate in the trial were asked to briefly explain reason(s) for their decision and to complete a questionnaire assessing their health-related quality of life, exercise behavior and attitudes towards exercise. Patients who participated in the PACES trial completed the same questionnaire as part of baseline assessments. All patients from whom data were collected provided written informed consent. The institutional review boards of all participating hospitals approved the study. Study measures Reasons for non-participation Reasons for non-participation were assessed with preset response options and an other category. Responses were aggregated to form two broad categories: (a) wanting to exercise on one s own; and (b) not wishing to exercise because it would be too burdensome practically, physically, or emotionally. Based on cognitive behavioral theory, we expected that factors associated significantly with the decision not to participate would be different for these two groups of decliners. Sociodemographic and clinical data Sociodemographic data, including age, sex, education, marital status, living and work situation, lifestyle variables, and comorbidity data, were obtained via the questionnaire. Comorbidity was assessed with a standard list of conditions used frequently in the Netherlands, including heart disease, respiratory disease, arthritis, chronic back problems, diabetes, diseases of the digestive tract, migraine, urinary tract disease, psychiatric problems, chronic skin conditions, and diseases of the central nervous system. Clinical information, including cancer stage, type of surgery, and radiotherapy, was obtained from the medical records. 98

Non-participation Fatigue, health-related quality of life, and psychological distress Fatigue was assessed with the Multidimensional Fatigue Inventory (MFI). 14 The 20-item MFI is organized into five scales: general fatigue, physical fatigue, reduced activity, reduced motivation, and mental fatigue. Health-related quality of life was assessed with the 2-item global quality of life scale of the European Organisation for Research and Treatment of Cancer Quality-of-Life Questionnaire (EORTC QLQ- C30). 15 Psychological distress was assessed with the total scale of the 14-item Hospital Anxiety and Depression Scale (HADS). 16 Cronbach s α of the QLQ-C30 global quality of life scale, the five MFI scales, and the HADS were 0.83, 0.88, 0.85, 0.87, 0.76, 0.89, and 0.90, respectively. Behavioral and attitudinal variables Current physical activity level was measured with the Physical Activity Scale for the Elderly (PASE). 17 The PASE consists of questions on leisure time, household, and work-related physical activities. The total activity score is computed by multiplying the hours per week spent on each activity by item weights and summing over all activities. 18 Exercise stage was based on the transtheoretical model and assessed the willingness/ preparedness to be physically active. 19 Participants were categorized as being in the precontemplation (no intention to exercise), contemplation (intention to exercise), preparation (irregular exercise), action (started exercising 30 min a day in last six months), or maintenance phase (exercising 30 min a day for longer than six months). A series of questions based on the theory of planned behavior 10 was posed to assess physical activity related self-efficacy, perceived social support, and barriers to and benefits of physical activity. 10,20,21 Self-efficacy was measured with five items with a 0 10 response scale, expressing the perceived likelihood of exercising when tired, in a bad mood, when feeling pressed for time, when on holiday, or with bad weather. 19 The Cronbach s α for this scale was 0.85. A higher score indicates a stronger sense of self-efficacy. Social support for being physically active during chemotherapy was measured with seven items with a 5-point Likert-type response scale. Items assessed perceived support from partner, family, friends, colleagues, general practitioner, treating physician, and other patients with cancer. 20 The overall social support scale had a Cronbach s α of 0.94. A higher score indicates higher levels of perceived social support. Items on perceived barriers to and benefits of physical activity were selected from two existing questionnaires. 10,11 The 13 items addressing barriers included lack of money, time, energy, motivation, transportation, support for exercise, counseling about exercise, limited possibilities in the environment, health conditions, nausea, fatigue, pain, and work responsibilities. Items were summed to form a total scale, with a Cronbach s α of 0.84. A higher score represents more perceived barriers. The 12 items addressing perceived benefits of regular physical activity included improved health and reduced risk of disease, feeling better about oneself, improved fitness, improved daily functioning, weight loss, meeting new people, getting one s mind off cancer and its treatment, improved overall well-being, maintaining a normal lifestyle, coping with the stress 6 99

of cancer and treatment, gaining control over cancer and life, and recovering from treatment. All items were scored on a 5-point Likert-type scale. The Cronbach s α of this scale was 0.91. Higher scores indicate more perceived benefits. Attitude towards physical activity was measured using 7-point adjective rating response scales that assessed both evaluative (useful useless, harmful beneficial, wise foolish, and bad good) and affective (enjoyable unenjoyable, boring interesting, pleasant unpleasant, and easy hard) dimensions. 10 The attitude scale had a Cronbach s α of 0.95. Higher scores indicate a more positive attitude towards exercise. Statistical analysis Descriptive statistics were generated to characterize the study sample. Differences between trial participants and non-participants who completed the questionnaire were analyzed with chi-squared statistic for categorical variables and analysis of variance for continuous variables. Post hoc analyses were used to analyze the differences between trial participants and non-participants who wished to exercise and between trial participants and those nonparticipants who did not wish to exercise. All statistical tests were two-sided, with α set at 5%. Mean differences and 95% confidence intervals were accompanied by r-type effect sizes (ES). 22 Effect sizes of 0.1 were considered small, 0.3 moderate, and 0.5 large. 22 RESULTS During the study period, 524 eligible breast cancer patients were referred to the PACES trial, of whom 230 (44%) consented to participate. Of the 294 non-participants, 131 (45%) completed the full questionnaire, while the other 163 (55%) provided only basic sociodemographic data and reasons for non-participation. Self-reported reasons for non-participation The main reasons for non-participation were having too many things on one s mind (34%), wanting to exercise on one s own outside the context of the trial (24%), issues with the logistics and timing of the trial (specifically, not willing or able to undergo baseline measurements before start of chemotherapy; 22%), too great a time investment (19%), not wanting to exercise at all (18%), travel distance (12%), and not wanting to participate in a clinical trial (e.g., not wanting to be randomized, to have to complete questionnaires, or undergo physical performance tests; 11%) (Table 1). Reasons for non-participation differed between those who completed the nonparticipants questionnaire and those who only provided reasons for non-participation. Nonparticipants who completed the questionnaire were significantly more likely to report travel distance (22% versus 4%, P < 0.01) and time it would cost them to participate in the study (32% versus 9%, P < 0.01), and significantly less frequently reported the timing of the trial (12% versus 29%, P < 0.01) as reasons for non-participation (Table 1). Based on reasons for non-participation, we grouped the non-participants who completed 100

Non-participation Table 1. Self-reported reasons for non-participation in the PACES trial (N=294) Reasons All NP (N=294) NP (N=131) NPB (N=163) P Too many things on mind 100 (34%). 39 (30%) 61 (37%) 0.17 Want to exercise by him/herself 72 (24%) 41 (31%) 31 (19%) 0.02 Poor timing 64 (22%) 16 (12%) 48 (29%). <0.01 Too time-consuming 56 (19%) 42 (32%) 14 (9%).. <0.01 Does not want to exercise 52 (18%) 30 (23%) 22 (13%) 0.04 Travel distance 35 (12%) 29 (22%) 6 (4%). <0.01 Does not want to participate in RCT 33 (11%) 14 (11%) 19 (12%) 0.79 Unknown 4 (1%) 0 (0%) 4 (2%) 0.07 NOTE: Bold font indicates significant difference. Patients could provide multiple reasons for non-participation. Abbreviations: All NP, both non-participant groups combined; NP, non-participants who provided both reasons for non-participation and completed the questionnaire; NPB, non-participants who provided only basic sociodemographic information and reasons for non-participation; RCT, randomized control trial. the questionnaire into (a) those who wished to exercise by themselves (N=41); and (b) those who did not wish to exercise in the context of a trial (N=90). Sociodemographic and clinical variables The mean age of all eligible patients was 51 years; there were only two male patients (Table 2). There was no significant age difference between trial participants and non-participants. In the remainder of this paper, we will compare trial participants to those non-participants who completed the non-participant questionnaire. Education level was not significantly different between non-participants who wished to exercise on their own and trial participants. Compared with trial participants, those nonparticipants who did not want to exercise had a significantly lower level of education (55% versus 30%, P < 0.01). Both of the non-participant groups were significantly less likely to be employed full-time, as compared with trial participants (12% versus 27%, P = 0.02 and 10% versus 27%, P < 0.01, respectively). Compared with trial participants, non-participants who did not wish to exercise were significantly more likely to have had a mastectomy (23% versus 39%, P < 0.01). No statistically significant differences were observed between non-participants and participants in marital status, cancer stage, lymph node dissection, radiotherapy, or comorbidity (Table 2). 6 Fatigue, health-related quality of life, and psychological distress Those non-participants who wished to exercise on their own did not differ significantly from trial participants in fatigue level or global health-related quality of life. Compared with trial participants, those who did not wish to exercise reported significantly more general fatigue (P < 0.01, ES = 0.51) and mental fatigue (P = 0.04, ES = 0.25), reduced activity (P < 0.01, ES = 0.35), reduced motivation (P < 0.01, ES = 0.50), and lower global health-related quality of life (P < 0.01, ES = 0.50). There were no statistically significant differences in psychological distress (Table 3). 101

Table 2. Baseline sample sociodemographic and clinical characteristics TP NP-E NP-O NPB overall TP vs. NP-E TP vs. NP-O TP vs. NPB Characteristic N=230 N=41 N=90 N=163 P P P P Age mean years (SD) 50.7 (9.1) 52.5 (9.8) 51.0 (10.8) 51.9 (10.5) 0.36 0.17 0.52 0.16 Gender N female (%) 228 (99%) 41 (100%) 90 (100%) 163 (100%) 0.56 0.49 0.35 0.45 Marital status N (%) 0.94 0.75 0.93 Single/divorced/widowed 50 (22%) 8 (20%) 20 (22%) Married/living together 180 (78%) 33 (80%) 70 (78%) Education N (%) 0.02 0.14 <0.01 Primary/middle school 43 (19%) 12 (29%) 26 (29%) High school 61 (26%) 13 (32%) 37 (41%) College/university 126 (55%) 16 (39%) 27 (30%) Work N (%) <0.01 0.02 <0.01 Full-time 63 (27%) 5 (12%) 9 (10%) Part-time 95 (41%) 17 (41%) 38 (42%) Other* 72 (31%) 19 (46%) 43 (48%) Cancer stage N (%) 0.14 0.92 0.03 Stage I 12 (5%) 3 (7%) 4 (4%) Stage II 109 (47%) 17 (41%) 29 (32%) Stage III 109 (47%) 20 (49%) 55 (61%) Cancer treatment N (%) Mastectomy 52 (23%) 11 (27%) 35 (39%) 0.01 0.52 <0.01 Breast-conserving surgery 178 (77%) 29 (71%) 53 (59%) Sentinel lymph node dissection 212 (92%) 38 (93%) 82 (91%) 0.81 0.53 0.99 Axillary lymph node dissection 71 (31%) 15 (37%) 23 (26%) 0.39 0.43 0.35 Radiotherapy 180 (78%) 34 (83%) 66 (73%) 0.63 0.41 0.75 Comorbidity N (%) 125 (54%) 20 (49%) 50 (56%) 0.73 0.59 0.65 Current smoker N (%) 29 (13%) 4 (10%) 12 (13%) 0.84 0.61 0.86 NOTE: Bold font indicates significant difference. Abbreviations: SD, standard deviation; TP, trial participants; NP-E, non-participants who completed the questionnaire, who wanted to exercise by themselves; NP-O, non-participants who completed the questionnaire, who cannot or do not want to exercise; NPB, non-participants who provided basic sociodemographic information and reasons for non-participation. *Students, home-makers, retired, and unemployed. 102

Non-participation Table 3. Descriptive statistics of and group differences in fatigue, quality of life and behavioral and attitudinal factors TP and NP-E TP and NP-O TP (N=230) NP-E (N=41) NP-O (N=90) Mean (SD) Mean (SD) Mean (SD) P Mean difference (95% CI) ES* P Mean difference (95% CI) ES* P Measure General fatigue a 11.0 (4.1) 12.2 (4.7) 13.3 (4.7). <0.01-1.1 (-2.6 ; 0.3) 0.26 0.12-2.3 (-3.3 ; -1.2) 0.51. <0.01 Physical fatigue a 10.3 (4.0) 10.2 (4.6) 11.9 (4.8) 0.01 1.1 (-0.4 ; 2.5) 0.26 0.14-1.0 (-2.1 ; 0.1) 0.23 0.06 Reduced activity a 10.5 (4.2) 9.4 (3.8) 11.6 (4.9) 0.03 0.1 (-1.3 ; 1.6) 0.03 0.85-1.6 (-2.6 ; -0.5) 0.35. <0.01 Reduced motivation a 8.7 (3.5) 9.1 (3.4) 10.6 (4.2). <0.01-0.3 (-1.6 ; 0.9) 0.10 0.58-1.9 (-2.8 ; -1.0) 0.50. <0.01 Mental fatigue a 9.9 (4.5) 9.3 (4.0) 11.1 (4.3) 0.05 0.7 (-0.8 ; 2.1) 0.16 0.37-1.1 (-2.2 ; -0.1) 0.25 0.04 Health-related quality of life b 73.9 (15.5) 71.3 (18.3) 64.6 (22.0). <0.01 2.6 (-3.3 ; 8.5) 0.15 0.38 9.3 (5.0 ; 13.7) 0.50. <0.01 Psychological distress c 8.7 (6.1) 8.6 (5.6) 9.9 (6.7) 0.30 0.1 (-2.0 ; 2.2) 0.02 0.92-1.2 (-2.7 ; 0.4) 0.18 0.14 Self-reported activity level d 99.0 (75.6) 132.1 (82.2) 121.3 (89.6) 0.01-33.1 (-60.1 ; -6.2) 0.42 0.02-22.4 (-42.2 ; -2.5) 0.27 0.03 Exercise stage N(%) 0.02 0.04 0.11 Pre-contemplation 4 (1.7%) 1 (2.4%) 3 (3.3%) Contemplation 48 (20.9%) 1 (2.4%) 17 (18.9%) Preparation 51 (22.2%) 11 (26.8%) 31 (34.4%) Action 8 (3.5%) 0 (0.0%) 1 (1.1%) Maintenance 119 (51.7%) 28 (68.3%) 37 (41.1%) Self-efficacy 6.6 (1.7) 6.3 (1.8) 5.8 (1.8). <0.01 0.3 (-0.3 ; 0.9) 0.16 0.33 0.8 (0.3 ; 1.2) 0.44. <0.01 Social support 4.5 (0.6) 4.3 (0.9) 4.1 (0.8). <0.01 0.2 (0.0 ; 0.5) 0.32 0.04 0.4 (0.2 ; 0.5) 0.52. <0.01 Benefits 3.9 (0.7) 3.8 (0.8) 3.7 (0.5) 0.01 0.1 (-0.1 ; 0.3) 0.11 0.46 0.3 (0.1 ; 0.4) 0.42. <0.01 Barriers 2.1 (0.5) 1.9 (0.4) 2.2 (0.9) 0.01 0.2 (0.0 ; 0.4) 0.39 0.10-0.2 (-0.3 ; 0.0) 0.26 0.02 Attitude 5.8 (1.1) 5.3 (1.5) 4.9 (1.5). <0.01 0.5 (0.1 ; 0.9) 0.36 0.03 0.9 (0.6 ; 1.2) 0.67. <0.01 NOTE: Bold font indicates significant difference. Abbreviations: SD, standard deviation; TP, trial participants; NP-E, non-participants who completed the questionnaire, who want to exercise by themselves; NP-O, non- participants who completed the questionnaire, who cannot or do not want to exercise. *ES, r-type effect size. a MFI, Multidemensional Fatigue Inventory (score range from 4 to 20, high scores indicating more fatigue). b EORTC CQL-C30 (score range from 0 to 100, higher scores indicating better global health status). c HADS, Hospital Anxiety and Depression Scale (score range from 0 to 42, higher score represents higher levels of psychological distress). d PASE: Physical Activity Scale for the Elderly (total activity score). 6 103

Behavior and attitude towards exercise during chemotherapy Non-participants who wanted to exercise on their own had significantly higher physical activity levels compared with trial participants (on average, 33 min per week more active (95% CI: 6 to 60, ES = 0.42, P = 0.02)). Compared with trial participants, non-participants who did not wish to exercise were, on average, 22 min per week more active (95% CI: 3 to 42, ES = 0.27, P = 0.03). Non-participants who wanted to exercise on their own were more likely to be in the maintenance exercise phase compared with trial participants (P = 0.04) (Table 3). Compared with trial participants, non-participants who did not wish to exercise had significantly lower levels of self-efficacy (P < 0.01, ES = 0.44), experienced less social support (P < 0.01, ES = 0.52), less benefits (P < 0.01, ES = 0.42), and more barriers (P = 0.02, ES = 0.26). Both non-participant groups had a more negative attitude towards exercise during chemotherapy than trial participants (P = 0.03, ES = 0.36 and P < 0.01, ES = 0.67, respectively) (Table 3). DISCUSSION In this study, we examined reasons for and sociodemographic and attitudinal factors associated with (non-)participation in an exercise trial during adjuvant chemotherapy for breast cancer. Slightly more than half of the patients eligible for the PACES trial declined participation. This non-participation rate is similar to that reported in previous exerciseoncology trials. 7,23,24 Patients who chose to participate in the trial were higher educated and more often employed than non-participants. This is a common finding in trials of psychosocial and behavioral interventions among cancer patients and other patient populations. 5 7,23,25 In general, higher education is associated positively with health literacy. Those who are health literate may be more open to advice about and more motivated to be physically active. 26 This might have led to selection of a more highly motivated group of patients who would be expected to perform better during the trial than the larger target population of interest. Conversely, if these patients were more physically fit at baseline, they might have had to exercise more intensely to maintain the same level of physical fitness. Thus, it is unclear whether selection of more highly educated patients might result in overestimation or underestimation of the effect of exercise. About one-third of the non-participants reported that they wished to exercise on their own. Not surprisingly, this group of non-participants reported a higher exercise stage and higher physical activity levels. Activity levels before diagnosis are predictive of activity levels post-treatment. 8 This suggests that these relatively active patients may not need or benefit from a formal, structured exercise program during treatment. As expected, a large group of non-participants did not wish to exercise in the context of a trial. These patients were significantly more likely to have had a mastectomy, be more fatigued, and report lower levels of health-related quality of life. One previous study 7 in healthy, inactive adults also reported that those with poorer health were less willing to participate in a physical activity promotion trial. In a feasibility study in patients with 104

Non-participation late stage non-small cell lung cancer, patients with more symptoms were less likely to participate. 27 Non-participants who did not wish to exercise experienced less social and professional support and had lower levels of self-efficacy in relation to exercise. They also perceived fewer benefits from and more barriers to being physically active. The importance of social support, perceived benefits, and perceived barriers to participation in physical activity programs has been reported in earlier studies with COPD and cerebral palsy patients. 28,29 It is possible that patients with more health problems are not encouraged, if not actively discouraged to exercise, by their caregivers and social network members. 30 Future research is needed to better understand how those patients with compromised health status, with attitudes and beliefs that are not conducive to exercising, and with social networks that do not encourage them to be physically active can nevertheless be motivated to participate in formal, supervised exercise during treatment. Where that is not feasible, our goal should be to at least increase their physical activity levels to match public health guidelines. Some of the patients who did not wish to exercise in the context of a trial reported barriers that were specific to the trial itself. Such trial-specific issues have been reported in other randomized controlled studies. 31,32 The design of the PACES trial required that patients were randomized before the start of their chemotherapy. For some patients, the emotional burden and logistical issues surrounding the start of treatment represented a substantial barrier to participation. In clinical practice, patients could enter an exercise program at any time during their treatment, which may enhance uptake. Similarly, in clinical practice, the burden of having to complete questionnaires and performance tests could be minimized. Finally, although we established a network of physical therapists from which patients could choose, we could not ensure close proximity to home for all patients. Thus, as the network of physical therapists and exercise physiologists trained in supervising cancer patients under active treatment increases, convenience should become a less relevant barrier to participation. People are known to over-report their physical activity levels. 33 While non-participants only provided self-reported physical activity levels, physical activity levels of trial participants were also objectively measured by accelerometry. Because participants knew they were being monitored, they may have reported their physical activity levels more accurately than the non-participants. This may explain, in part, why higher physical activity levels were reported by both non-participant groups. The current study had several limitations. First, although the percentage of nonparticipants who completed our questionnaire was substantial (slightly under half), the results cannot necessarily be generalized to the whole group of non-participants. In fact, those non-participants who did not complete our questionnaire were more likely to report practical barriers (e.g., too time-consuming, travel distance, and poor timing) than those nonparticipants who completed the questionnaire. Second, the only data available to compare trial participants with non-participants were attitudes and self-reported behavior. It would be instructive to have performance-based information about non-participants to determine the extent to which their actual physical 6 105

fitness affected their decision. Our study also had a number of notable strengths, including extensive data on reasons for non-participation and detailed information on sociodemographic, clinical, attitudinal, and behavioral factors potentially associated with non-participation. Summarizing, our results indicate that patients with breast cancer who chose not to participate in a physical exercise trial during their adjuvant chemotherapy treatment fall roughly into two categories. The first was those who wished to exercise on their own and thus did not feel the need to follow a formal program or did not want to risk randomization to a control group. In theory, they should not be of too great concern to us, although it is important to note that this was the self-reported reason for non-participation. We did not have any objective data on actual physical activity levels or physical fitness with which to confirm their self-report data. The second group included those who did not wish to exercise in the context of a trial. This group deserves additional attention. We have identified a number of potentially important attitudinal, contextual, and practical barriers that are actionable. This includes using new recruitment and educational strategies that can be adapted to different stages of readiness to be physically active. It is also important to educate both informal and formal caregivers about the potential benefits of exercising during treatment, so that they can provide appropriate encouragement and support. 106

Non-participation REFERENCES 1. Puetz TW, Herring MP. Differential effects of exercise on cancer-related fatigue during and following treatment: A meta-analysis. Am J Prev Med, 2012; 43: e1-24. 2. Jones LW, Liang Y, Pituskin EN, et al. Effect of exercise training on peak oxygen consumption in patients with cancer: A meta-analysis. Oncologist, 2011; 16: 112 20. 3. Mishra SI, Scherer RW, Snyder C, et al. Exercise interventions on health-related quality of life for people with cancer during active treatment. Cochrane database Syst Rev, 2012; 8: CD008465. 4. Chang B-H, Hendricks AM, Slawsky MT, et al. Patient recruitment to a randomized clinical trial of behavioral therapy for chronic heart failure. BMC Med Res Methodol, 2004; 4: 8. 5. Lloyd-Williams F, Mair F, Shiels C, et al. Why are patients in clinical trials of heart failure not like those we see in everyday practice? J Clin Epidemiol, 2003; 56: 1157 62. 6. Würtzen H, Dalton SO, Andersen KK, et al. Who participates in a randomized trial of mindfulness-based stress reduction (MBSR) after breast cancer? A study of factors associated with enrollment among Danish breast cancer patients. Psycho-oncology, 2013; 22: 1180 5. 7. Chinn DJ, White M, Howel D, et al. Factors associated with non-participation in a physical activity promotion trial. Public Health, 2006; 120: 309 19. 8. Craike M, Hose K, Livingston PM. Physical activity participation and barriers for people with multiple myeloma. Support Care Cancer, 2013; 21: 927 34. 9. van Waart H, Stuiver MM, van Harten WH, et al. Effect of low-intensity physical activity and moderateto high-intensity physical exercise during adjuvant chemotherapy on physical fitness, fatigue, and chemotherapy completion rates: Results of the PACES randomized clinical trial. J Clin Oncol, 2015; 33: 1918 27. 10. Courneya KS, Friedenreich CM. Utility of the theory of planned behavior for understanding exercise during breast cancer treatment. Psycho-oncology, 1999; 8: 112 22. 11. van der Ploeg HP, van der Beek AJ, van der Woude LH V, et al. Physical activity for people with a disability: A conceptual model. Sports Med, 2004; 34: 639 49. 12. Borg G. Borg s percieved exertion and pain scales. 1998th ed. Champaign, IL: Human Kinetic; 2010. 13. de Backer IC, Schep G, Hoogeveen A, et al. Exercise testing and training in a cancer rehabilitation program: The advantage of the steep ramp test. Arch Phys Med Rehabil, 2007; 88: 610 6. 14. Smets EM, Garssen B, Bonke B, et al. The multidimensional fatigue inventory (MFI) psychometric qualities of an instrument to assess fatigue. J Psychosom Res, 1995; 39: 315 25. 15. Aaronson NK, Ahmedzai S, Bergman B, et al. The European organization for research and treatment of cancer QLQ-C30: A quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst, 1993; 85: 365 76. 16. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand, 1983; 67: 361 70. 17. Liu RDK, Buffart LM, Kersten MJ, et al. Psychometric properties of two physical activity questionnaires, the AQuAA and the PASE, in cancer patients. BMC Med Res Methodol, 2011; 11: 30. 18. Washburn RA, McAuley E, Katula J, et al. The physical activity scale for the elderly (PASE): evidence for validity. J Clin Epidemiol, 1999; 52: 643 51. 19. Marcus BH, Selby VC, Niaura RS, et al. Self-efficacy and the stages of exercise behavior change. Res Q Exerc Sport, 1992; 63: 60 6. 20. Sallis JF, Calfas KJ, Alcaraz JE, et al. Potential mediators of change in a physical activity promotion course for university students: Project GRAD. Ann Behav Med, 1999; 21: 149 58. 21. Biddle S, Goudas M, Page A. Social-psychological predictors of self-reported actual and intended physical activity in a university workforce sample. Br J Sports Med, 1994; 28: 160 3. 22. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale NJ: Lawrence Erlbaum; 1988. 6 107

23. Sears SR, Stanton AL, Kwan L, et al. Recruitment and retention challenges in breast cancer survivorship research: Results from a multisite, randomized intervention trial in women with early stage breast cancer. Cancer Epidemiol Biomarkers Prev, 2003; 12: 1087 90. 24. Courneya KS, McKenzie DC, Mackey JR, et al. Effects of exercise dose and type during breast cancer chemotherapy: Multicenter randomized trial. J Natl Cancer Inst, 2013; 105: 1821 32. 25. Lakerveld J, Ijzelenberg W, van Tulder MW, et al. Motives for (not) participating in a lifestyle intervention trial. BMC Med Res Methodol, 2008; 8: 17. 26. von Wagner C, Knight K, Steptoe A, et al. Functional health literacy and health-promoting behaviour in a national sample of British adults. J Epidemiol Community Health, 2007; 61: 1086 90. 27. Cheville AL, Girardi J, Clark MM, et al. Therapeutic exercise during outpatient radiation therapy for advanced cancer: Feasibility and impact on physical well-being. Am J Phys Med Rehabil, 2010; 89: 611 9. 28. Thorpe O, Johnston K, Kumar S. Barriers and enablers to physical activity participation in patients with COPD: A systematic review. J Cardiopulm Rehabil Prev, 2012; 32: 359 69. 29. Verschuren O, Wiart L, Hermans D, et al. Identification of facilitators and barriers to physical activity in children and adolescents with cerebral palsy. J Pediatr, 2012; 161: 488 94. 30. Barber FD. Social support and physical activity engagement by cancer survivors. Clin J Oncol Nurs, 2012; 16: E84-98. 31. Prescott RJ, Counsell CE, Gillespie WJ, et al. Factors that limit the quality, number and progress of randomised controlled trials. Health Technol Assess, 1999; 3: 1 143. 32. Ross S, Grant A, Counsell C, et al. Barriers to participation in randomised controlled trials: a systematic review. J Clin Epidemiol, 1999; 52: 1143 56. 33. Adams SA. The Effect of Social Desirability and Social Approval on Self-Reports of Physical Activity. Am J Epidemiol, 2005; 161: 389 98. 108