Behaviour change techniques in physical activity interventions for men with prostate cancer: A systematic review

Abstract

Physical activity interventions can improve prostate cancer survivors’ health. Determining the behaviour change techniques used in physical activity interventions can help elucidate the mechanisms by which an intervention successfully changes behaviour. The purpose of this systematic review was to identify and evaluate behaviour change techniques in physical activity interventions for prostate cancer survivors. A total of 7 databases were searched and 15 studies were retained. The studies included a mean 6.87 behaviour change techniques (range = 3–10), and similar behaviour change techniques were implemented in all studies. Consideration of how behaviour change techniques are implemented may help identify how behaviour change techniques enhance physical activity interventions for prostate cancer survivors.

Keywords

exercise behaviour health behaviour physical activity prostate cancer systematic review

Background

In Canada, one in seven men is diagnosed with prostate cancer during his lifetime, making prostate cancer the most commonly diagnosed type of cancer among men (Canadian Cancer Statistics, 2017). Since 2001, the rate of death from prostate cancer has steadily declined by 3.3 per cent each year due to improved screening and treatment options, with a 5-year survival rate of 95 per cent (Canadian Cancer Statistics, 2017). As a result of more cancer survivors¹ living longer, more men are also living with consequences of treatment, such as fatigue and poorer health-related quality of life (Bourke et al., 2013). Declines in physical and psychological health can prevent prostate cancer survivors from engaging in physical activity and increase their risk of developing comorbidities, such as cardiovascular disease (Keogh and MacLeod, 2012).

Physical activity helps to improve management and recovery from cancer (Pedersen and Saltin, 2015) and has been shown to be safe and feasible for cancer survivors (Schmitz et al., 2010). The American College of Sports Medicine (ACSM) acknowledges that cancer survivors can benefit from physical activity and should obtain at least 150 minutes of moderate-intensity aerobic activity and 2–3 days of strength training per week (Schmitz et al., 2010). Despite the ACSM guidelines, up to 70 per cent of prostate cancer survivors are not sufficiently active (Thorsen et al., 2008). Cancer survivors may have added difficulty being active as a result of symptoms associated with treatment and could benefit from support through behaviour change interventions (Ferrer et al., 2011). Therefore, several physical activity behaviour change interventions have been developed to improve physical activity levels and overall health of prostate cancer survivors.

Reviews of physical activity and exercise interventions among prostate cancer survivors show that aerobic and resistance training are effective at improving: aerobic fitness, muscular strength, physical functioning, quality of life and fatigue (Baumann et al., 2012; Bourke et al., 2016; Keogh and MacLeod, 2012). There is strong evidence that exercise interventions can have positive psychological and physical benefits for prostate cancer survivors; however, these results must be interpreted with caution. There is large heterogeneity among interventions in terms of support provided to participants, outcomes reported, duration, frequency, type of exercise and setting (Baumann et al., 2012). The variability in the implementation process of these components in physical activity interventions can have a notable impact on intervention success (Hilton and Johnston, 2017). For example, the quality of the interaction between interventionists and participants can impact the success of an intervention whereby a higher quality relationship is more likely to lead to a more effective intervention (Horvath et al., 2011). There is also considerable inconsistency across interventions in terms of reported outcomes. In some studies, the impact of the intervention is determined by increases in physical activity, whereas other studies report physical and/or psychological health improvements that are attributed to increases in physical activity. Reviews of physical activity interventions by Baumann et al. (2012) and Keogh and MacLeod (2012) have demonstrated that supervised or group-based interventions had greater health improvements compared to unsupervised or home-based programmes. Some studies showed resistance training improved quality of life, fitness and fatigue more so than aerobic training (Baumann et al., 2012; Keogh and MacLeod, 2012). Therefore, it appears certain intervention components may be more effective than others at improving health outcomes and changing physical activity levels among prostate cancer survivors.

Physical activity behaviour change interventions typically involve numerous interacting components, making it difficult to determine which active components make them effective interventions (Michie et al., 2013). However, it is essential to identify these components to create more effective interventions. Michie et al. (2013) developed a taxonomy of active components called behaviour change techniques (BCTs). A BCT is a distinct strategy that helps change or adjust the processes that govern behaviour (Michie et al., 2011). The mechanisms by which physical activity interventions among prostate cancer survivors improve health outcomes are only speculative (Bourke et al., 2016). Identifying the active components, or BCTs, in interventions can help elucidate the mechanisms and effects of physical activity interventions for prostate cancer survivors and provide insight into what makes some interventions more successful at changing behaviour than others. Therefore, the primary purpose of this systematic review was to identify BCTs used in physical activity behaviour change interventions for prostate cancer survivors. The secondary purpose was to evaluate and determine the potential contribution of BCTs to increase physical activity levels among prostate cancer survivors.

Methods

Methodological framework

A systematic review of the literature was conducted using the preferred reporting items for systematic reviews and meta-analyses (PRISMA) checklist (Supplemental material 1).

Study inclusion criteria

Types of studies

Studies were included if they were published in English in peer-reviewed journals. Study designs included randomized controlled trials, quasi-experimental studies and pre-experimental (one group pre-test/post-test design) studies. Protocol papers associated with studies meeting the inclusion criteria were included. Protocol papers that did not have published results were excluded.

Types of participants

Participants in eligible studies were men diagnosed with prostate cancer (i.e. not yet receiving treatment, undergoing treatment or post-treatment). Studies that targeted any other population, such as other types of cancer were excluded. Studies that invited caregivers to attend were included only if the focus of the intervention was on prostate cancer survivors and separate data were available for the prostate cancer survivors.

Types of interventions

The studies included supervised or unsupervised physical activity or exercise interventions. Physical activity can include exercise, as well as other activities such as household chores and active transport. Studies that had interventions focused on multiple behaviours, such as diet and physical activity, were excluded. The physical activity interventions required the presence of at least one BCT according to the BCT Taxonomy (v1; Michie et al., 2013).

Types of outcome measures

The main outcome of interest was measures of physical activity (e.g. steps per day, minutes of physical activity and physical activity frequency). Given the heterogeneity in outcomes reported, studies were included that did not report a direct measure of physical activity, but measured outcomes hypothesized to result from increases in physical activity.

Data sources

A systematic literature search was conducted using the following databases: MEDLINE (OVID), PsycINFO (OVID), EMBASE (OVID), CINAHL Plus, SPORTDiscus, Cochrane Library and PubMed for published articles (from 1990 to 2016). The search was performed in July 2016. Relevant search words were categorized into two distinct themes: (a) population and (b) physical activity. The search terms were developed in MEDLINE (OVID) and adapted for the remaining databases (as shown in Supplemental Material 2). The reference lists of full-text articles selected and relevant systematic reviews were scanned for additional articles. The registered trial number of full-text articles was searched in order to find any other associated articles. All references were downloaded into the EndNote X7 reference management software package for assessment.

Data collection

Selection of studies

The search strategy and eligibility criteria were discussed by two authors. The literature search was conducted by one author who removed duplicates and screened titles for eligibility. Two authors independently screened the abstracts, and three authors independently screened any relevant full-text articles. Throughout the process of article screening and selection, any disagreements regarding eligibility were discussed until a consensus was reached.

Data extraction

Data were extracted independently by two authors following the data extraction checklist from the Cochrane Handbook for Systematic Reviews (Higgins and Green, 2011). This checklist included authors, year of publication, country, study design, number of participants, mean age, intervention description, study duration, attrition, adherence and physical activity outcomes (if available). Participants’ medical information, such as stage of cancer and treatment type and duration, were extracted at the level of detail as described in the studies (i.e. not all medical information was available for each study). The data extracted were confirmed by a third author, and any discrepancies were discussed until an agreement was reached.

Intervention classification system

The studies were classified based on the potential of the intervention to increase physical activity levels using a system developed by Gardner et al. (2016). Interventions were categorized as ‘very promising’, ‘quite promising’, ‘non-promising’ or ‘unclear promise’ based on within-group and between-group analyses showing statistically significant increases in physical activity. Interventions were considered very promising if there was a significant increase in at least one measure of physical activity within the intervention group, and the increase was statistically greater than the change observed in at least one comparator group (e.g. control group). Interventions were classified as quite promising if there was a significant increase in at least one measure of physical activity within the intervention group or the increase was statistically greater than the change observed in at least one comparator group. If an intervention did not include a comparator arm but showed promise, it was classified as quite promising. Interventions were categorized as non-promising if there was no statistically significant increase in physical activity within the intervention group, nor statistical differences relative to a comparator group (Gardner et al., 2016). Interventions that did not report an explicit measure of physical activity were categorized as unclear promise.

BCT coding

The physical activity interventions were coded using the BCT Taxonomy (v1) which defines 93 BCTs and is the standard for identifying and classifying BCTs for health behaviour change. All authors completed an online training (Wood et al., 2015), specifically developed for this taxonomy, which enhanced the consistency and reliability of BCT coding. The target behaviour of the BCTs was physical activity and the target population was prostate cancer survivors. If a BCT appeared in a protocol paper, it was surmised that the BCT was implemented during intervention delivery and in the article reporting the results. The first four interventions were coded independently, and the authors came together to compare, discuss and clarify any ambiguity between codes. A consensus process was used to resolve any disagreements by referring back to the taxonomy and a final decision was made. The authors came together to discuss the codes after every five articles. This method resulted in consistent coding among authors and in the end, resulted in 100 per cent agreement for all BCTs coded. Where insufficient details were given in a manuscript, the excerpt from the text was not coded to avoid making an assumption that a BCT was used.

BCT analysis

In order to judge the potential contribution of BCTs used in an intervention, a ‘promise ratio’ was calculated for specific BCTs (Gardner et al., 2016; Martin et al., 2013) by dividing the number of very and quite promising interventions that implemented the BCT by the number of non-promising interventions that implemented that same BCT. A BCT was considered promising if it was used in at least twice as many promising interventions as non-promising interventions (i.e. promise ratio ≥2) and was found in at least two promising interventions. If a BCT only appeared in promising interventions, the number of interventions in which it appeared was reported (Gardner et al., 2016). A promise ratio was not calculated for BCTs that only appeared in non-promising interventions or only appeared once to avoid making conclusions with insufficient evidence. One-way analysis of variance (ANOVA) was conducted to determine the difference in mean BCTs implemented in the different promise categories.

Quality assessment

The methodological quality and risk of bias were assessed for each individual study using a combination of criteria from the Cochrane Collaboration’s Risk of Bias tool (Higgins and Green, 2011) and criteria proposed by Chambless and Hollon (1998), similarly used by Lambert et al. (2016). The quality assessment included the following criteria: (a) sufficient power for a moderate effect, (b) inclusion criteria, (c) reliability or validity of outcome measures reported, (d) random sequence generation, (e) allocation concealment, (f) blinding of personnel, (g) blinding assessment outcome, (h) incomplete data reported and (i) selective reporting. The articles were assessed independently by two authors. Prior to coding, the definitions of the criteria were discussed to ensure the criteria were all assessed in the same manner. Each domain was marked with ‘yes’ (value of 1) or ‘no’ (value of 0). For quasi-experimental and pre-experimental studies, where certain criteria did not apply, they were marked with ‘not applicable (N/A)’ and given a value of 0 to reflect the higher potential for bias in such studies. Where some information was reported but not all, such as for reliability or validity of outcome measures reported, it was rated with a ‘no’. Studies were deemed to be high quality if they received a score of 7–9, moderate quality with a score of 4–6 and low quality with a score of 0–3.

Results

The search yielded 9493 titles from the databases, which was reduced to 6971 titles after duplicates were removed. A screening of titles and abstracts resulted in 82 full-text articles to be reviewed. Of those 82 full text-articles, 15 studies were retained (with four additional papers including three protocol papers and one secondary analysis; Supplemental material 3). The results are reported on data extracted and coded from all the articles but presented as 15 studies.

Overview of included studies

Table 1 describes the 15 studies included in this review. Most studies were conducted in Canada (n = 7), with others in the United States (n = 3), Australia (n = 3), Sweden (n = 1) and Spain (n = 1). The mean ages of participants ranged from 64.5 to 77.5 years. The number of participants in each study ranged from 13 to 859. The participants had various stages of prostate cancer, received a wide range of treatments and differing treatment durations. The types of treatments included watchful waiting, androgen deprivation therapy (ADT) or hormone therapy, brachytherapy, radiation therapy, chemotherapy, surgery, prostatectomy or a combination of several treatments. The study designs included randomized controlled trials (n = 10), quasi-experimental designs (n = 2) and pre-experimental designs (n = 3).

Table 1.

Characteristics of included studies.

Study	Population, setting	Conditions/assessments	Outcomes (measures)
Berglund et al. (2007) N = 211 RCT	Mean age (years): 69 Treatment: watchful waiting: 30%, radical prostatectomy: 26%, radiation: 9%, hormone therapy: 26% and unknown: 9% Country: Sweden	T1 (n = 53): 60 minutes, seven sessions, light physical training, relaxation and breathing exercises, booster session 2 months after training T2 (n = 55): 60 minutes, seven sessions and information about prostate cancer T3 (n = 52): 135 minutes, seven sessions, light physical training, relaxation and breathing exercises + information about prostate cancer. Booster session 2 months after training C (n = 51): standard care provided by hospital Duration: unclear Ax: Pre, 6 months FU and 12 months FU	Attrition: 25% Adherence: not reported No PA outcomes reported
Beydoun et al. (2014) N = 859 Quasi-experimental design (non-equivalent control group design)	Mean age (years): 71 Treatment: all receiving hormone therapy Country: Australia	T1 (n = 255): 6 months, home-based exercise video and graded resistance bands + telephone support from health coach at regular intervals T2 (n = 396): 2×/week for 10 weeks, supervised group sessions, followed by 6 months home-based exercise or education about health C (n = 208): education about health Duration: 10 weeks Ax: Pre, post 10 weeks (T2 only), 3 months FU and 6 months FU. Primary outcomes assessed at 10 weeks only	Attrition: 8.7% Adherence: 90% for home-based exercise (not reported for supervised sessions) No PA outcomes reported
Campo et al. (2014) N = 40 RCT	Mean age (years): 72 Median time since diagnosis: 5.6 years Treatment: ADT: 48%, radical prostatectomy: 45%, surgery: 10%, radiation: 24%, brachytherapy: 17% and chemotherapy: 7% Stage I–II: 34% Stage III–IV: 56% Country: United States	T (n = 16): 60 minutes, 2×/week, 12 weeks, instructor led Qigong classes and supplemented with home-based Qigong practice C (n = 13): active control, light stretching exercises but avoided relaxation or meditation techniques Duration: 12 weeks Ax: Pre and 1 week post (13 weeks)	Attrition: 20% for T and 35% for C Adherence: 85% for T and 43% for C T = C: home-based PA frequency
Carmack Taylor et al. (2004, 2006) N = 134 RCT	Mean age (years): 69.2 Time on ADT: <21 months: 50% >21 months: 50% Country: United States	T1 (n = 46): 90 minutes, 16 weekly sessions and 4 bimonthly sessions, PA training sessions and information about behavioural skills for PA T2 (n = 51): 90 minutes, 16 weekly sessions and 4 bimonthly sessions, information sessions about prostate cancer C (n = 37): information on resources for prostate cancer survivors, usual care Duration: 6 months Ax: Pre, post 6 months and 12 months FU	Attrition: 17% at 6 months and 16% at 12 months Adherence: 64% (13.4/21) sessions for T1% and 70% (14.8/21) for T2 T1 = T2 = C: days/week active 30 + minutes and energy expenditure
Culos-Reed et al. (2007) N = 31 Pre-experimental design (one group with pre-test and post-test)	Mean age (years): 64.8 Time since diagnosis (mean): 33.9 months Treatment: planning to receive ADT for at least 6 months Country: Canada	T: 12 weeks, home-based, walking, stretching, light resistance activities 3–5×/week. Booster sessions (group-based) for PA and education: 90 minutes, every 2 weeks (in weeks 1, 3, 5, 7, 9 and 11) C: N/A Duration: 3 months Ax: Pre, post 3 months and 7 months FU	Attrition: 42% attrition for 4 months follow-up questionnaire Adherence: 81% attended 5 or 6 of 6 booster sessions Pre = post: mild Pre < post: moderate and strenuous PA frequency, total PA frequency and leisure score index Post = follow-up: mild and moderate PA frequency, total PA frequency and leisure score index Post < follow-up: strenuous PA frequency
Culos-Reed et al. (2010) N = 100 RCT	Mean age (years): T = 67.2 C = 68 Treatment: all receiving ADT Country: Canada	T (n = 53): 3–5×/week, 16 weeks, home-based, walking, stretching, light resistance exercises, moderate intensity and 3–5×/week. Booster sessions (group-based) for PA and education, 90 minutes, weekly for 16 weeks and then monthly sessions during FU C (n = 47): wait-list control Duration: 16 weeks Ax: Baseline, post 16 weeks, 6 months FU and 10 months FU (no data reported for 6 months and 10 months FU)	Attrition: 34% Adherence: 77.8% T > C: increase in leisure score index
Galvão et al. (2009, 2014) N = 100 RCT	Mean age (years): T = 71.9 C = 71.5 Time since diagnosis (mean): 5.6 years Previous ADT duration (mean): 12 months Stage II: 62% Stage III–IV: 38% Country: Australia	T (n = 50): 2×/week, 6 months, vigorous aerobic and resistance training, supervised in exercise clinic, supplemented with 2×/week (recommended) home-based aerobic sessions. Followed by 6 months home-based resistance, aerobic and flexibility exercises. C (n = 50): received pedometer and educational booklet with recommendations to do PA for 12 months Duration: 6 months Ax: Pre, post 6 months and 12 months FU	Attrition: 16% at 6 months and 12% at 12 months for T, 10% at 6 months and 6% at 12 months for C Adherence: 77% T = C: steps per day (pedometer) or self-reported PA at 6 months and 12 months
Livingston et al. (2011, 2015),^a and Gaskin et al. (2016) N = 147 RCT	Mean age (years): T = 66.9 C = 64.7 Time since treatment (mean): 25 weeks Treatment: surgery: 44%, surgery + radiation: 21%, surgery + radiation + ADT: 6%, radiation: 14%, radiation + ADT: 14% and surgery + ADT: 1% Stage I: 42% Stage II: 39% Stage III: 19% Country: Australia	T (n = 54): 50 minutes, 2×/week,12 weeks, aerobic, resistance and flexibility exercises, moderate to vigorous intensity, supervised at a local gym and supplemented with one weekly home-based session C (n = 93): usual care (brief advice on exercise training) Duration: 12 weeks Ax: Pre and post 12 weeks	Attrition: 13% for T and 11% for C Adherence: 85% adhered to at least 18 of 24 sessions for T T = C: self-reported MVPA minutes/week and moderate-intensity PA minutes/week; accelerometer data for minutes/day, MVPA, vigorous and moderate intensity T > C: self-reported vigorous intensity activity minutes/week; percentage meeting MVPA guidelines per week
McGowan et al. (2013) N = 423 RCT	Mean age (years): T1 = 67.9 T2 = 68.4 T3 = 68.9 Treatment: watchful waiting: 27.7%, surgery 48.5%, radiation: 44.9%, chemotherapy: 2.8%, hormone therapy: 32.9% and cancer reoccurrence: 3.1% In treatment: 68.8% completed treatment: 18.2% Country: Canada	T1 (n = 141): PA guideline fact sheet with recommendations T2 (n = 141): self-administrated implementation intention of techniques to change PA behaviour + PA fact sheet T3 (n = 141): self-administered implementation intention of techniques to change PA behaviour + one telephone-assisted session for goal setting and planning + PA fact sheet Duration: 1 session + phone call Ax: Pre, 1 months and 3 months FU	Attrition: not reported for PA Adherence: not reported for PA 1 month: T2 > T1: PA minutes/week and moderate PA minutes/week T1 = T2: vigorous PA minutes/week T1 = T3: PA minutes/week, moderate PA minutes/week and vigorous PA minutes/week T2 = T3: PA minutes/week, moderate PA minutes/week and vigorous PA minutes/week 3 months: T1 = T2 = T3: PA minutes/week, moderate PA minutes/week and vigorous PA minutes/week
Sajid et al. (2016) N = 19 RCT	Mean age (years): T1 = 75.7 T2 = 77.5 C = 71.8 Time on ADT (mean): 62 months Country: United States	T1 (n = 6): 6 week, home-based, moderate-intensity aerobic (walking) and progressive resistance exercise programme. Weekly phone call reminders. Followed by 6 weeks without reminders T2 (n = 8): 6 week, home-based exercise programme with Wii-Fit. Weekly phone call reminders. Followed by 6 weeks without reminders C (n = 5): usual care Duration: 6 weeks Ax: Pre, post 6 weeks and 12 week FU	Attrition: 38% for T1, 17% for T2% and 40% for C Adherence: not reported T1 > C: increase in steps per day (pedometer)
Santa Mina et al. (2012) N = 13 RCT	Mean age (years): T1 = 70.5 T2 = 66.3 Treatment: on LHRH: 69%, LHRH + ADT: 8%, discontinued ADT in the past 3 months: 23% Country: Canada	T1 (n = 6): 60 minutes, 3×/week, group-based sessions, resistance and vigorous aerobic exercises T2 (n = 7): 60 minutes, 3×/week, personal trainer guided sessions, resistance and vigorous aerobic exercises Duration: 8 weeks Ax: Pre and post 8 weeks	Attrition: 50% for T1% and 0% for T2 Adherence: 88% for T1% and 91% for T2 No PA outcomes reported
Santa Mina et al. (2013) N = 66 RCT	Mean age (years): T1 = 72.1 T2 = 70.6 Treatment: LHRH: 57.5%, LHRH + ADT: 35% and ADT alone: 7.5% Stage I: 15% Stage II: 35% Stage III: 21% Stage IV: 15% Unknown: 14% Country: Canada	T1 (n = 32): 30–60 minutes, 3–5×/week, 6 months, home-based moderate-to-vigorous intensity aerobic exercise programme. Booster sessions (group-based), 90 minutes and every 2 weeks (12 sessions total) T2 (n = 34): 30–60 minutes, 3–5×/week, 6 months and home-based moderate-to-vigorous intensity resistance exercise programme. Booster sessions (group-based), 90 minutes and every 2 weeks (12 sessions total) Duration: 6 months Ax: Pre, mid 3 months, post 6 months and 12 months FU	Attrition: 26% for T1% and 44% for T2 at 6 months Adherence: T1 attended mean 16.4% of booster sessions and T2 attended mean 5.6% of booster sessions T1 > T2: PA volume (MET-hour/week) at 3 months, 6 months and 12 months
Serda et al. (2010) N = 36 Pre-experimental design (one group with pre-test and post-test)	Mean age (years): 71.78 Treatment: surgery: 45.5%, ADT: 45.5%, radiation + ADT: 3% and prostatectomy + ADT: 6% Stage II: 39.4% Stage III: 54.6% Stage IV: 3% Unknown: 3% Country: Spain	T: 90 minutes, 2×/week, 16 weeks, resistance training, supervised and moderate intensity. Followed by 8 weeks of unsupervised exercise C: N/A Duration: 24 weeks Ax: Pre and post 24 weeks	Attrition: 8% Adherence: 93% No PA outcomes reported
Truong et al. (2011) N = 80 Quasi-experimental (non-equivalent control group design)	Mean age (years): T1 = 67 C = 69 ADT duration (mean): 12 months Treatment: No ADT: 7.5%, LHRH: 11% and total androgen blockade: 81.5% Stage I: 9% Stage II: 48% Stage III: 33% Unknown: 9% Country: Canada	T (n = 50): 20 minutes, 3×/week and moderate-intensity walking, C (n = 30): historical comparison group with usual care Duration: 12 weeks Ax: Pre, mid 3–4 weeks, post 6–8 weeks and 6 months FU	Attrition: 16% for T Adherence: 88% No PA results reported
Zahavich et al. (2013) N = 15 Pre-experimental (one group with pre-test and post-test)	Mean age (years): 64.5 Treatment: ADT: 26.7%, no ADT: 73.3%, surgery: 53.3% and no surgery: 46.7% Post-treatment: 13.3% In treatment: 6.7% Country: Canada	T: 75 minutes, 7 weeks, yoga sessions led by instructor. Followed by 7 week self-selected PA Duration: 7 weeks Ax: Pre, post 7 weeks and 14 weeks FU	Attrition: 7% at 14 week FU Adherence: 87% Pre = post: PA duration frequency or leisure score index total score

RCT: randomized controlled trial; ADT: androgen deprivation therapy; LHRH: luteinizing hormone–releasing hormone; T: treatment condition; C: comparator condition; Duration: length of intervention; Ax: timing of assessments; Attrition: the percentage of participants without follow-up data; Adherence: as explicitly reported by authors; T > C: treatment significantly superior to control; T = C: no significant differences between groups; PA: physical activity; MVPA: moderate-to-vigorous physical activity; FU: follow-up; MET: metabolic equivalent.

Data extracted and reported from Livingston et al. (2015).

Types of interventions

The types of physical activity performed in the interventions varied among the 15 studies. The activities included aerobic training (n = 1; Truong et al., 2011), resistance training (n = 1; Serda et al., 2010), combined aerobic and resistance exercises (n = 8; Beydoun et al., 2014; Culos-Reed et al., 2007, 2010; Galvão et al., 2014; Livingston et al., 2015; Sajid et al., 2016; Santa Mina et al., 2012, 2013), yoga (n = 1; Zahavich et al., 2013), Qigong practices (n = 1; Campo et al., 2014) and unspecified (n = 1; Berglund et al., 2007). Two studies (Carmack Taylor et al., 2006; McGowan et al., 2013) focused on discussing physical activity and teaching participants BCTs, such as goal setting, action planning and self-monitoring of behaviour. The studies were conducted either in a gym or exercise clinic (n = 1), were strictly home-based (n = 3) or were both clinic and home-based (n = 11). All of the interventions implemented in clinics were supervised by an accredited exercise physiologist or kinesiologist. Interventions lasted from 2 to 12 months,² with most interventions encouraging weekly physical activity. Seven interventions offered telephone support or booster sessions in order to increase adherence to the programme. Nine studies had control groups that were subjected to usual care, received information regarding physical activity and health or practised light stretching.

Intervention classification

The 15 studies were classified into four categories based on the promise of the intervention to change physical activity behaviour. Five interventions (33.33%; Culos-Reed et al., 2010; Livingston et al., 2015; McGowan et al., 2013; Sajid et al., 2016; Santa Mina et al., 2013) were classified as very promising, one intervention (6.67%; Culos-Reed et al., 2007) as quite promising and four interventions (26.67%; Campo et al., 2014; Carmack Taylor et al., 2006; Galvão et al., 2014; Zahavich et al., 2013) as non-promising. Five interventions (33.33%; Berglund et al., 2007; Beydoun et al., 2014; Santa Mina et al., 2012; Serda et al., 2010; Truong et al., 2011) had unclear promise as they did not include a direct measure of physical activity. Analyses were conducted by grouping very promising and quite promising interventions to create an all-encompassing promising category.

Physical activity outcomes

The physical activity outcomes of the six promising and four non-promising interventions included self-reported and objective measures (using pedometers and accelerometers), such as steps per day, minutes of physical activity and leisure score index from the Godin Leisure Time Exercise Questionnaire (Godin and Shephard, 1985). The six promising studies (Culos-Reed et al., 2007, 2010; Livingston et al., 2015; McGowan et al., 2013; Sajid et al., 2016; Santa Mina et al., 2013) showed significant increases in physical activity levels post-intervention. Of those six, Santa Mina et al. (2013) and Sajid et al. (2016) showed significant increases in physical activity levels at follow-up (at 6 and 12 months and 12 weeks, respectively). The four non-promising studies (Campo et al., 2014; Carmack Taylor et al., 2006; Galvão et al., 2014; Zahavich et al., 2013) showed nonsignificant increases in any measures of physical activity levels post-intervention.

BCTs

The BCTs identified in each study are summarized in Table 2. The interventions included as few as three BCTs (Serda et al., 2010) to as many as 10 BCTs (Livingston et al., 2015). There was a mean of 6.87 BCTs per study, with a median of 7 and mode of 5. Of the possible 93 BCTs, 21 (22.58%) were present in at least one intervention. Behavioural practice/rehearsal is the only BCT that appeared in all interventions (n = 15). The other most common BCTs included instruction on how to perform the behaviour (n = 11), self-monitoring of behaviour (n = 9), adding objects to the environment (n = 9), social support (unspecified; n = 8), generalization of target behaviour (n = 8) and graded tasks (n = 7).

Table 2.

BCTs coded in the 15 studies based on the BCT Taxonomy (v1).

	Berglund et al. (2007)	Beydoun et al. (2014)	Campo et al. (2014)	Carmack Taylor et al. (2006)	Culos-Reed et al. (2007)	Culos-Reed et al. (2010)	Galvão et al. (2014)	Livingston et al. (2015)	McGowan et al. (2013)	Sajid et al. (2016)	Santa Mina et al. (2012)	Santa Mina et al. (2013)	Serda et al. (2010)	Truong et al. (2011)	Zahavich et al. (2013)
1.1 Goal setting (behaviour)				X					X
1.2 Problem solving				X					X
1.4 Action planning									X
1.5 Review behaviour goals				X
2.2 Feedback on behaviour				X
2.3 Self-monitoring of behaviour			X	X	X		X	X		X		X		X	X
2.6 Biofeedback							X				X	X		X
3.1 Social support (unspecified)		X	X	X	X	X		X				X			X
3.2 Social support (practical)		X						X	X
4.1 Instruction on how to perform the behaviour	X	X	X				X	X	X	X	X	X		X	X
4.2 Information about antecedents					X	X		X
5.3 Information about social and environmental consequences		X					X		X		X	X
6.1 Demonstration of the behaviour	X	X	X	X								X			X
7.1 Prompts and cues	X								X	X
8.1 Behavioural practice/rehearsal	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X
8.6 Generalization of target behaviour		X	X	X			X	X		X			X		X
8.7 Graded tasks			X				X	X		X		X	X		X
9.1 Credible source							X	X	X
10.2 Material reward (behaviour)											X
10.3 Non-specific reward				X
12.5 Adding objects to the environment		X	X		X	X	X	X		X		X		X
Total BCTs	4	8	8	10	5	4	9	10	9	7	5	9	3	5	7

BCT: behaviour change technique.

BCTs per intervention classification

The number of BCTs implemented in promising interventions (M = 7.33, standard deviation (SD) = 2.42), non-promising interventions (M = 8.50, SD = 1.29) and interventions with unclear promise (M = 5.00, SD = 1.87) was not statistically different (F(2,14) = 3.65, p = 0.058). The number of BCTs implemented in promising interventions did not significantly differ from the number of BCTs implemented in non-promising interventions (F(1,9) = 0.761, p = 0.408). A promise ratio was calculated for 16 BCTs, of which 6 BCTs were deemed promising (Table 3). The six promising BCTs included social support (practical; ratio = 2), information about antecedents (ratio = 3), information about social and environmental consequences (ratio = 2), prompts and cues (ratio = 2), credible source (ratio = 2) and adding objects to the environment (ratio = 2.5). Social support (practical), information about antecedents and prompts and cues appeared in promising interventions and not in non-promising interventions. Information about antecedents is the one BCT that appeared only in promising interventions and not in interventions classified as non-promising or unclear promise.

Table 3.

BCTs implemented per intervention classification and promise ratios.

	Types of interventions					Promise ratio
	Very promising (n = 5)	Quite promising (n = 1)	Non-promising (n = 4)	Unclear promise (n = 5)	All (n = 15)	Promise ratio
1.1 Goal setting (behaviour)	1	0	1	0	2	1
1.2 Problem solving	1	0	1	0	2	1
1.4 Action planning	1	0	0	0	1	–
1.5 Review behaviour goals	0	0	1	0	1	–
2.2 Feedback on behaviour	0	0	1	0	1	–
2.3 Self-monitoring of behaviour	3	1	4	1	9	1
2.6 Biofeedback	1	0	1	2	4	1
3.1 Social support (unspecified)	3	1	3	1	8	1.33
3.2 Social support (practical)	2	0	0	1	3	2
4.1 Instruction on how to performthe behaviour	4	0	3	4	11	1.33
4.2 Information about antecedents	2	1	0	0	3	3
5.3 Information about social and environmental consequences	2	0	1	2	5	2
6.1 Demonstration of the behaviour	1	0	3	2	6	0.33
7.1 Prompts and cues	2	0	0	1	3	2
8.1 Behavioural practice/rehearsal	5	1	4	5	15	1.5
8.6 Generalization of target behaviour	2	0	4	2	8	0.5
8.7 Graded tasks	3	0	3	1	7	1
9.1 Credible source	2	0	1	0	3	2
10.2 Material reward (behaviour)	0	0	0	1	1	–
10.3 Non-specific reward	0	0	1	0	1	–
12.5 Adding objects to the environment	4	1	2	2	9	2.5

BCT: behaviour change technique.

Rows in bold indicate promising BCTs given they have a promise ratio ≥2 and found in at least two interventions.

Methodological quality

The overall quality of the 15 studies is found in Supplemental material 4. Only one article (Carmack Taylor et al., 2006) was deemed high quality. Eight studies were of moderate quality (Berglund et al., 2007; Campo et al., 2014; Galvão et al., 2014; Livingston et al., 2015; McGowan et al., 2013; Sajid et al., 2016; Santa Mina et al., 2012, 2013) and six studies were of low quality (Beydoun et al., 2014; Culos-Reed et al., 2007, 2010; Serda et al., 2010; Truong et al., 2011; Zahavich et al., 2013). In total, 10 studies were underpowered for a moderate effect. All studies except for one (Beydoun et al., 2014) specified inclusion criteria and only one study (Campo et al., 2014) reported the validity and reliability of all measures. For randomized controlled trials, randomization, allocation concealment and blinding were typically addressed and reported. In all 15 studies, incomplete data were explicitly addressed. Four articles (Culos-Reed et al., 2007, 2010; Sajid et al., 2016; Serda et al., 2010) failed to report results for all of the outcomes measured (i.e. selective reporting).

Discussion

The primary purpose of this systematic review was to identify BCTs implemented in interventions aimed at increasing physical activity among prostate cancer survivors. The secondary purpose was to evaluate BCTs by determining their potential contribution to change behaviour and increase physical activity. The common BCTs identified were behavioural practice/rehearsal, instruction on how to perform the behaviour, self-monitoring of behaviour, adding objects to the environment, social support (unspecified) and generalization of target behaviour. The BCTs identified in this review are very similar to BCTs identified in physical activity interventions among other populations, such as adults with rheumatoid arthritis (Larkin et al., 2015), older adults (French et al., 2014) and cancer survivors (Bourke et al., 2013, 2016). Given that these interventions are all focused on changing physical activity behaviour, the consistency in BCTs is not surprising.

The studies that were considered promising had very similar BCTs as those in the non-promising interventions. A similar observation is made in a review of physical activity behaviour change interventions among adults with rheumatoid arthritis (Larkin et al., 2015). In our review, the non-promising interventions even had slightly more BCTs (8.50 BCTs compared to 7.33 BCTs, on average) than the promising interventions, although not significantly different. In line with previous research and recommendations by Michie et al. (2009), our findings suggest that the number of BCTs may not be the best indicator of what makes one intervention more successful than another. Including fewer BCTs that are delivered well may be more effective than a larger quantity of BCTs with a suboptimal delivery. Therefore, determining the dose, the frequency or the quality of BCT delivery may be necessary to explain the circumstances under which BCTs effectively change behaviour.

In this review, six studies were considered promising interventions to change physical activity behaviour. Among these six studies, various BCTs were either actively discussed or taught to the participants by those delivering the intervention. For example, in a study by McGowan et al. (2013), participants were instructed how to set S.M.A.R.T. goals, action plan and problem solve. One intervention group further received telephone support from a counsellor experienced in delivering physical activity behaviour change interventions to assist participants in creating realistic goals and plans (McGowan et al., 2013). The educational component within promising interventions may highlight an important component to help increase the potential for behaviour change. When BCTs are not taught, survivors may rely too much on those delivering the intervention for support and fail to use these techniques post-intervention in order to change behaviour or sustain the behaviour change.

Four interventions were deemed non-promising, meaning they did not demonstrate a big potential to increase physical activity levels. One of the non-promising studies implemented an educational behaviour change component; however, the participants did not significantly increase physical activity levels (Carmack Taylor et al., 2006). The results of the study indicated participants mastered the techniques they learned and increased their stage of change. However, an underpowered sample size and a measure not sensitive enough to capture small changes in physical activity can perhaps explain the lack of behaviour change (Carmack Taylor et al., 2006). The remaining three non-promising studies did not indicate BCTs were taught to the participants and how to properly use them to facilitate behaviour change. Failing to teach participants how to use BCTs could explain why no significant changes in physical activity levels were observed in the three studies.

Social support (practical), prompts and cues, and information about antecedents were the only three BCTs implemented in promising interventions and not in non-promising interventions. Social support (practical) was coded for interventions in supervised exercise settings or exercise programmes monitored by exercise physiologists or kinesiologists. Perhaps supervised exercise has more promise to increase physical activity compared to unsupervised or home-based training where practical support may not be readily accessible. Prompts and cues was used through booster sessions (Berglund et al., 2007), telephone calls (Livingston et al., 2015) or reminders on the fridge (McGowan et al., 2013) to prompt participants to be active. A study by Phillips and Gardner (2016) demonstrated that prompts and cues for exercise was a predictor of exercise frequency and can be an important technique to increase engagement in physical activity. Information about antecedents was the one BCT that only appeared in promising interventions (and not in any of the remaining categories). This BCT involves making individuals aware of situations, contexts or emotions that can predict engagement in physical activity. If prostate cancer survivors can identify their personal antecedents that facilitate and predict physical activity engagement, it can lead to higher levels of physical activity. Michie et al. (2014: 151) noted that information about antecedents has been used less frequently in educational interventions; however, there appears to be some value in it to help participants change their behaviour.

Limitations and future directions

A significant problem with behaviour change interventions is they are seldom described with sufficient detail. Therefore, it is possible that BCTs were implemented but not adequately described to be coded. Three articles in this review were published protocol papers describing the methods. The assumption was made that if a BCT appeared in a protocol, it was also present in the article with the published results. Michie et al. (2009) suggest quality and delivery can play important roles in the effective implementation and contribution of a BCT. Despite the benefit of quantifying the quality and delivery of the BCTs, this was not conducted in the current review because the dose and frequency of the BCTs in most cases were reported with insufficient detail. This limitation has been acknowledged in similar reviews of physical activity interventions among other populations (Larkin et al., 2015; Michie et al., 2009). Hilton and Johnston (2017) have noted that BCTs provide an indication of what components may be helpful for changing behaviour; however, reporting the where, when, why, who and how the BCTs were implemented has been largely ignored in research articles. As a result, there is currently a lack of empirically based recommendations as to the method and style of delivery of BCTs that are likely to encourage physical activity behaviour change and sustained physical activity. Intervention protocols and studies must provide a more comprehensive report of the BCTs used by explicitly describing what BCTs were used but more importantly, how, why, when and where each BCT was implemented. Future studies could then examine the dose, frequency and quality of BCTs implemented in behaviour change interventions. The BCTs with the most promise could be identified and provide recommendations for designing future physical activity interventions and optimize behaviour change.

Given the heterogeneity in outcomes reported for the included physical activity interventions, there was not one consistent physical activity measure across all 15 studies. The statistical significance of the change in physical activity levels was the most common outcome reported across studies. However, only 10 of the 15 studies included in this review reported a measure of physical activity levels to provide a direct indication of behaviour change. These five studies were classified as having an unclear promise and a promise ratio could not be calculated for the BCTs. To further assess the potential of physical activity interventions and BCTs, a clearly defined measure of physical activity should be reported and more information should be provided to more accurately determine the degree to which an intervention successfully or unsuccessfully changes behaviour.

In order to determine the potential contribution of interventions and BCTs to change behaviour, the promise of interventions and BCTs were calculated. As noted in the study by Gardner et al. (2016), the classification system was ultimately arbitrary and may have led to the under- or over-estimation of the potential contribution of certain BCTs. For example, the potential of the intervention by Culos-Reed et al. (2007) could have been underestimated given it was classified as quite promising because it did not include a comparator arm, but if it had been tested against a comparator, it may have proven to be more effective and classified as very promising. Gardner et al. (2016) also stated the criteria used to calculate and determine the promise ratio may have been too conservative, which may underestimate the potential of BCTs not used as frequently. It may not be that the BCT is ‘ineffective’ but that there is simply not enough evidence to support its potential for behaviour change. However, due to the heterogeneity in study design, interventions and reported outcomes, using the classification system and promise ratios provides an appropriate method to synthesize and analyse the BCTs.

Conclusion

Physical activity is documented to help with management and recovery from cancer treatment (Schmitz et al., 2010), and prostate cancer survivors should be encouraged to engage in physical activity. This is the first review to identify and evaluate the potential contribution of BCTs in physical activity interventions among prostate cancer survivors. As evidenced by this review, the number and types of BCTs implemented in promising and non-promising interventions did not significantly differ. This highlights the need for future studies that assess the dose, frequency and quality of BCT delivery in order to provide more evidence of which BCTs have the most potential to increase physical activity levels among men with prostate cancer. There is evidence in this review to suggest BCTs that are properly taught and explained to cancer survivors might be a promising approach to increase physical activity levels. Therefore, prostate cancer survivors may respond better to programmes or healthcare providers who go beyond just providing them with information about physical activity but instruct survivors how and why to use BCTs as helpful tools to change physical activity behaviour.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

Supplemental material

Supplemental material is available for this article online.

Notes

ORCID iD

Laura Hallward

References

Baumann

Zopf

Bloch

(2012) Clinical exercise interventions in prostate cancer patients: A systematic review of randomized controlled trials. Supportive Care in Cancer 20: 221–233.

Berglund

Petersson

Eriksson

, et al. (2007) ‘Between men’: A psychosocial rehabilitation programme for men with prostate cancer. Acta Oncologica 46: 83–89.

Beydoun

Bucci

Chin

, et al. (2014) Prospective study of exercise intervention in prostate cancer patients on androgen deprivation therapy. Journal of Medical Imaging and Radiation Oncology 58: 369–376.

Bourke

Homer

Thaha

, et al. (2013) Interventions for promoting habitual exercise in people living with and beyond cancer. Cochrane Database of Systematic Reviews 9: CD010192.

Bourke

Smith

Steed

, et al. (2016) Exercise for men with prostate cancer: A systematic review and meta-analysis. European Urology 69: 693–703.

Campo

Agarwal

LaStayo

, et al. (2014) Levels of fatigue and distress in senior prostate cancer survivors enrolled in a 12-week randomized controlled trial of Qigong. Journal of Cancer Survivorship: Research and Practice 8: 60–69.

Canadian Cancer Statistics (2017). Cancer Society’s Advisory Committee on Cancer Statistics. Available at: http://www.cancer.ca/en/cancer-information/cancer-101/canadian-cancer-statistics-publication/?region=on-ixzz4edPEjYln.

Carmack Taylor

Smith

Moor

, et al. (2004) Quality of life intervention for prostate cancer patients: Design and baseline characteristics of the active for life after cancer trial. Controlled Clinical Trials 25: 265–285.

Carmack Taylor

Demoor

Smith

, et al. (2006) Active for life after cancer: A randomized trial examining a lifestyle physical activity program for prostate cancer patients. Psycho-Oncology 15: 847–862.

10.

Chambless

Hollon

(1998) Defining empirically supported therapies. Journal of Consulting and Clinical Psychology 66: 7–18.

11.

Culos-Reed

Robinson

Lau

, et al. (2007) Benefits of a physical activity intervention for men with prostate cancer. Journal of Sport & Exercise Psychology 29: 118–127.

12.

Culos-Reed

Robinson

Lau

, et al. (2010) Physical activity for men receiving androgen deprivation therapy for prostate cancer: Benefits from a 16-week intervention. Supportive Care in Cancer 18: 591–599.

13.

Ferrer

Huedo-Medina

Johnson

, et al. (2011) Exercise interventions for cancer survivors: A meta-analysis of quality of life outcomes. Annals of Behavioral Medicine 41: 32–47.

14.

French

Olander

Chisholm

, et al. (2014) Which behaviour change techniques are most effective at increasing older adults self-efficacy and physical activity behaviour? A systematic review. Annals of Behavioral Medicine 48: 225–234.

15.

Galvão

Spry

Taaffe

, et al. (2009) A randomized controlled trial of an exercise intervention targeting cardiovascular and metabolic risk factors for prostate cancer patients from the RADAR trial. BMC Cancer 9: 419.

16.

Galvão

Spry

Denham

, et al. (2014) A multicentre year-long randomised controlled trial of exercise training targeting physical functioning in men with prostate cancer previously treated with androgen suppression and radiation from TROG 03.04 RADAR. European Urology 65: 856–864.

17.

Gardner

Smith

Lorencatto

, et al. (2016) How to reduce sitting time? A review of behaviour change strategies used in sedentary behaviour reduction interventions among adults. Health Psychology Review 10: 89–112.

18.

Gaskin

Fraser

Owen

, et al. (2016) Fitness outcomes from a randomised controlled trial of exercise training for men with prostate cancer: the ENGAGE study. Journal of Cancer Survivorship 10: 972–980.

19.

Godin

Shephard

(1985) A simple method to assess exercise behavior in the community. Canadian Journal of Applied Sport Sciences 10: 141–146.

20.

Higgins

JPT

Green

(2011) Cochrane Handbook for Systematic Reviews of Interventions (Version 5.1.0). The Cochrane Collaborationg. Available at: http://handbook.cochrane.org

21.

Hilton

Johnston

(2017) Health psychology: It’s not what you do, it’s the way that you do it. Health Psychology Open 4: 1–10.

22.

Horvath

Del Re

Fluckiger

, et al. (2011) Alliance in individual psychotherapy. Psychotherapy 48: 9–16.

23.

Keogh

MacLeod

(2012) Body composition, physical fitness, functional performance, quality of life, and fatigue benefits of exercise for prostate cancer patients: A systematic review. Journal of Pain and Symptom Management 43: 96–110.

24.

Lambert

Duncan

Kapellas

, et al. (2016) A descriptive systematic review of physical activity interventions for caregivers: Effects on caregivers and care recipients psychosocial outcomes, physical activity levels, and physical health. Annals of Behavioral Medicine 50: 907–919.

25.

Larkin

Gallagher

Cramp

, et al. (2015) Behaviour change interventions to promote physical activity in rheumatoid arthritis: A systematic review. Rheumatology International 35: 1631–1640.

26.

Livingston

Salmon

Courneya

, et al. (2011) Efficacy of a referral and physical activity program for survivors of prostate cancer [ENGAGE]: Rationale and design for a cluster randomised controlled trial. BMC Cancer 11: 237.

27.

Livingston

Mohebbi

Craike

, et al. (2015) Effects of a clinician referral and exercise program for men who have completed active treatment for prostate cancer: A multicenter cluster randomized controlled trial (ENGAGE). Cancer 121: 2646–2654.

28.

McGowan

North

Courneya

(2013) Randomized controlled trial of a behavior change intervention to increase physical activity and quality of life in prostate cancer survivors. Annals of Behavioral Medicine 46: 382–393.

29.

Martin

Chater

Lorencatto

(2013) Effective behaviour change techniques in the prevention and management of childhood obesity. International Journal of Obesity 37: 1287–1294.

30.

Michie

Abraham

Eccles

, et al. (2011) Strengthening evaluation and implementation by specifying components of behaviour change interventions: A study protocol. Implementation Science 6: 10.

31.

Michie

Abraham

Whittington

, et al. (2009) Effective techniques in healthy eating and physical activity interventions: A meta-regression. Health Psychology 28: 690–701.

32.

Michie

Atkins

West

(2014) The Behaviour Change Wheel: A Guide to Designing Interventions. London: Silverback Publishing.

33.

Michie

Richardson

Johnston

, et al. (2013) The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: Building an international consensus for the reporting of behavior change interventions. Annals of Behavioral Medicine 46: 81–95.

34.

National Coalition for Cancer Survivorship (2014) Defining cancer survivorship. Available at: https://www.canceradvocacy.org/news/defining-cancer-survivorship/

35.

Pedersen

Saltin

(2015) Exercise as medicine – Evidence for prescribing exercise as therapy in 26 different chronic diseases. Scandinavian Journal of Medicine & Science in Sports 25(Suppl. 3): 1–72.

36.

Phillips

Gardner

(2016) Habitual exercise instigation (vs. execution) predicts healthy adults’ exercise frequency. Health Psychology 35: 69–77.

37.

Sajid

Dale

Mustian

, et al. (2016) Novel physical activity interventions for older patients with prostate cancer on hormone therapy: A pilot randomized study. Journal of Geriatric Oncology 7: 71–80.

38.

Santa Mina

Alibhai

Matthew

, et al. (2013) A randomized trial of aerobic versus resistance exercise in prostate cancer survivors. Journal of Aging and Physical Activity 21: 455–478.

39.

Santa Mina

Ritvo

Matthew

, et al. (2012) Group exercise versus personal training for prostate cancer patients: A pilot randomized trial. Journal of Cancer Therapy 3: 146–156.

40.

Schmitz

Courneya

Matthews

, et al. (2010) American College of Sports Medicine roundtable on exercise guidelines for cancer survivors. Medicine and Science in Sports and Exercise 42: 1409–1426.

41.

Serda

Monreal

Del Valle

(2010) Physical exercise as complementary treatment in prostate cancer. Apunts Medicina De L’esport 45: 81–93.

42.

Thorsen

Courneya

Stevinson

, et al. (2008) A systematic review of physical activity in prostate cancer survivors: Outcomes, prevalence, and determinants. Supportive Care in Cancer 16: 987–997.

43.

Truong

Gaul

McDonald

, et al. (2011) Prospective evaluation of a 12-week walking exercise program and its effect on fatigue in prostate cancer patients undergoing radical external beam radiotherapy. American Journal of Clinical Oncology 34: 350–355.

44.

Wood

Richardson

Johnston

, et al. (2015) Applying the behaviour change technique (BCT) taxonomy v1: A study of coder training. Translational Behavioral Medicine 5: 134–148.

45.

Zahavich

ANR

Robinson

Paskevich

, et al. (2013) Examining a therapeutic yoga program for prostate cancer survivors. Integrative Cancer Therapies 12: 113–125.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.45 MB