Skip to main content

Effect of self-regulatory behaviour change techniques and predictors of physical activity maintenance in cancer survivors: a 12-month follow-up of the Phys-Can RCT

Abstract

Background

Current knowledge about the promotion of long-term physical activity (PA) maintenance in cancer survivors is limited. The aims of this study were to 1) determine the effect of self-regulatory BCTs on long-term PA maintenance, and 2) identify predictors of long-term PA maintenance in cancer survivors 12 months after participating in a six-month exercise intervention during cancer treatment.

Methods

In a multicentre study with a 2 × 2 factorial design, the Phys-Can RCT, 577 participants with curable breast, colorectal or prostate cancer and starting their cancer treatment, were randomized to high intensity exercise with or without self-regulatory behaviour change techniques (BCTs; e.g. goal-setting and self-monitoring) or low-to-moderate intensity exercise with or without self-regulatory BCTs. Participants’ level of PA was assessed at the end of the exercise intervention and 12 months later (i.e. 12-month follow-up), using a PA monitor and a PA diary. Participants were categorized as either maintainers (change in minutes/week of aerobic PA ≥ 0 and/or change in number of sessions/week of resistance training ≥0) or non-maintainers. Data on potential predictors were collected at baseline and at the end of the exercise intervention. Multiple logistic regression analyses were performed to answer both research questions.

Results

A total of 301 participants (52%) completed the data assessments. A main effect of BCTs on PA maintenance was found (OR = 1.80, 95%CI [1.05–3.08]) at 12-month follow-up. Participants reporting higher health-related quality-of-life (HRQoL) (OR = 1.03, 95%CI [1.00–1.06] and higher exercise motivation (OR = 1.02, 95%CI [1.00–1.04]) at baseline were more likely to maintain PA levels at 12-month follow-up. Participants with higher exercise expectations (OR = 0.88, 95%CI [0.78–0.99]) and a history of tobacco use at baseline (OR = 0.43, 95%CI [0.21–0.86]) were less likely to maintain PA levels at 12-month follow-up. Finally, participants with greater BMI increases over the course of the exercise intervention (OR = 0.63, 95%CI [0.44–0.90]) were less likely to maintain their PA levels at 12-month follow-up.

Conclusions

Self-regulatory BCTs improved PA maintenance at 12-month follow-up and can be recommended to cancer survivors for long-term PA maintenance. Such support should be considered especially for patients with low HRQoL, low exercise motivation, high exercise expectations or with a history of tobacco use at the start of their cancer treatment, as well as for those gaining weight during their treatment. However, more experimental studies are needed to investigate the efficacy of individual or combinations of BCTs in broader clinical populations.

Trial registration

NCT02473003 (10/10/2014).

Peer Review reports

Background

There is strong evidence that regular physical activity (PA) improves many cancer-related health outcomes such as cancer-related fatigue, depression, quality of life and physical function [1]. Furthermore, emerging observational data suggest that regular PA can reduce cancer recurrence [2, 3] and increase overall survival [4]. Although cancer survivors have much to gain from regular PA, the majority tend to reduce their level of PA after being diagnosed [5, 6]. A PA decline is usually seen during the years following a cancer diagnosis [5], which indicates that engaging in PA during and after a cancer treatment is challenging for many cancer survivors. Therefore, many interventions have been designed to promote PA in this population and short-term PA increases have been demonstrated [7, 8]. However, sustaining this behaviour change is crucial to achieve long-term health benefits. PA maintenance, defined as a sustained PA behaviour for at least six months after the end of an intervention [9, 10], is therefore an area of particular interest in exercise oncology but is still insufficiently studied [7, 8].

Some systematic reviews have suggested that cancer survivors may benefit from behavioural interventions promoting long-term PA [8, 11]. Such interventions typically involve strategies to facilitate PA, including the use of behaviour change techniques (BCTs), defined as specific components of an intervention designed to change a behaviour [12]. A taxonomy of 93 BCTs has been developed by Michie et al. [12] to provide a standardised method of classifying such intervention components. Using this taxonomy, the above-mentioned reviews [8, 11] have reported that most interventions, targeting long-term PA, use different combinations of BCTs. Graded tasks, instructions on how to perform a specific PA behaviour and social support are examples of commonly used BCTs in interventions involving cancer survivors, irrespective of their efficacy in promoting PA [8, 11]. On the other hand, self-regulatory BCTs such as action planning, goal-setting and self-monitoring are frequently used in interventions involving different adult populations and resulting in high PA maintenance [8, 13]. Such BCTs are likely to facilitate self-regulation (i.e. ability to control/regulate one’s behaviour), which is a central concept in several theoretical frameworks such as Social Cognitive Theory [14] and has been identified as an important skill for successful PA maintenance [13]. Consequently, self-regulatory BCTs may be useful to maintain PA but more research is needed to confirm their possible impact on long-term PA maintenance in cancer survivors. We recently reported the effect of self-regulatory BCTs (e.g. goal-setting and self-monitoring) on cancer-related fatigue [15] and exercise adherence [16] in patients undergoing curative cancer treatment and participating in a six-month exercise intervention, the Physical Training and Cancer randomised controlled trial (Phys-Can RCT). We found no effect of self-regulatory BCTs on these outcomes during the intervention [15, 16]. However, determinants of PA behaviour in the short term are not necessarily the same in the long term [13]. Therefore, further investigations regarding the effect of self-regulatory BCTs on long-term PA are needed.

Additionally, increased knowledge about who may benefit most from behaviour change support is needed to develop more effective exercise interventions and optimize the use of such support in cancer survivors. For this purpose, it has been recommended to use a wide biopsychosocial approach [17]. This approach recognises the influence of multiple factors in PA behaviour change, including biological, psychological and social factors, and emphasizes that those different factors need to be considered to fully understand a PA behaviour [17]. Therefore, identifying predictors of long-term PA maintenance, based on physical, psychological and sociodemographic characteristics, both at baseline and post-intervention, is crucial to target vulnerable patient groups early in the course of their cancer treatment [18, 19] and after participating in an exercise intervention [20, 21]. Several factors such as body mass index (BMI) [22] and cardiovascular fitness [21, 22] (physical factors), exercise motivation [21, 22], exercise expectancy [22, 23] and exercise self- efficacy [10, 24] (psychological factors) as well as age [10, 22] and education [18] (sociodemographic factors) have been identified as potential predictors of PA maintenance. However, study findings are mixed and thus inconclusive. Furthermore, the majority of studies have examined predictors of short-term maintenance (e.g. 6 months post-intervention) [25], limiting the possibility to draw conclusions about predictors of PA maintenance in the longer term (12 months or more).

To date, knowledge about long-term PA maintenance after an exercise intervention in cancer survivors is limited. The aims of this study were therefore to 1) determine the effect of self-regulatory BCTs on long-term PA maintenance, and 2) identify baseline and post-intervention predictors of long-term PA maintenance in cancer survivors 12 months after participating in an exercise intervention during cancer treatment.

Methods

Study design and participants

This study is a secondary analysis using data from the Phys-Can RCT. The Phys-Can RCT is a multicentre study with a 2 × 2 factorial design (NCT02473003), approved by the Regional Ethical Review Board in Uppsala (Dnr 2014/249) [15, 26]. The Phys-Can RCT was designed to compare the effects of six months of high (HI) versus low-to-moderate intensity (LMI) exercise with or without self-regulatory BCTs on cancer-related fatigue in patients undergoing curative cancer treatment. All participants provided written informed consent. The recruitment process has previously been described in detail [15]. Briefly, eligible patients newly diagnosed with breast, colorectal or prostate cancer and scheduled to undergo (neo-)adjuvant cancer treatment, were recruited at university hospitals in three cities in Sweden (Lund/Malmö, Linköping and Uppsala) between March 2015 and April 2018. Participants were randomised to one of four intervention groups: 1) HI with self- regulatory BCTs, 2) LMI with self-regulatory BCTs, 3) HI without self-regulatory BCTs or 4) LMI without self-regulatory BCTs.

Participants included in the present study were those with PA data available at the end of the exercise intervention (6 months after randomization) and at 12-month follow-up (12 months after the end of the exercise intervention). Further, participants were included if they met international exercise guidelines for cancer survivors [1, 27] at the end of exercise the intervention (i.e. at least 75 min/week of vigorous intensity aerobic PA or 150 min/week of moderate intensity aerobic PA or 90 min/week of moderate-to-vigorous intensity aerobic PA, and/or two sessions of resistance training/week).

Intervention

Exercise programme

The exercise programme has been described previously in detail [15]. Briefly, participants exercised for six months while undergoing cancer treatment. The exercise programme consisted of supervised group-based resistance training and individual home-based endurance training. The resistance training included six machine-based exercises and was performed twice a week. Participants alternated between 3 × 6 repetitions maximum (RM) and 3 × 10 RM in the HI groups, and 3 × 12 repetitions at 50% of 6 RM and 3 × 20 repetitions at 50% of 10 RM in the LMI groups. The endurance training consisted of twice-weekly interval training (20–40 min/session) at 80–90% of heart rate reserve (HRR) in the HI groups, and 150 min weekly continuous-based exercise at 40–50% of HRR in the LMI groups.

Self-regulatory BCTs

Participants randomised to receive self-regulatory BCTs were guided by study coaches to use goal-setting, review of behavioural goal, self-monitoring, action planning and problem solving as described in Table 1. These BCTs were provided face-to-face, jointly with the resistance training sessions on a maximum of nine occasions, except for self-monitoring which was performed by the participants after each exercise session. Weekly behavioural meetings were offered during the first month to provide participants with a gradual introduction to the use of the different BCTs. For example, the study coaches assisted participants in formulating concrete and realistic goals as well as exercise plans specifying when, where and how to perform endurance training that was home-based. After the first month, those meetings were held every 4–6 weeks, depending on individual needs. The study coaches used printed sheets during each meeting to make notes about what was decided (goal-setting, exercise planning). Those notes were then discussed during the next meeting, and if participants did not manage to exercise according to their planning, strategies to overcome barriers to exercise were identified and adjustments were made. Participants also developed an individual written plan for relapse prevention at the end of the exercise intervention, including goal-setting and coping planning. This plan was individually followed up and revised at three and nine months post-exercise intervention in face-to-face or telephone meetings with the study coaches (Table 1).

Table 1 Self-regulatory BCTs provided in the Phys-Can RCT

Assessment of physical activity

PA was assessed at the end of the exercise intervention (6 months after randomisation) and at 12-month follow-up (12 months after the end of the exercise intervention), using previously validated instruments: a tri-axial PA monitor, the SenseWear Armband mini (SWA) [28, 29] and a 7-day PA diary [30]. At each measurement point, participants were asked to wear the SWA 24 h a day for seven consecutive days and fill out the diary during the same period.

To assess the weekly aerobic PA duration and intensity, data from the SWA were analysed with the SenseWear Professional 8.1 Software. The software provided SWA wear time and minutes spent in different levels of Metabolic Equivalent Task values (METs). PA at moderate and vigorous intensity was determined using established cut-off values, i.e. 3.0–5.9 METs and ≥ 6.0 METs respectively [31]. To reflect one week of PA, data from the SWA were included in the analyses if the SWA was worn for at least four days, including one weekend day, with a wear time of at least 80% per day [32,33,34,35]. Weekly time spent in the different intensity levels was calculated by summing minutes in 24 h for each valid day that the criterion for the relevant intensity was met.

To assess the weekly frequency of resistance training, data from the 7-day PA diary were used [30]. Participants made daily notes regarding the PA they performed, including type of activities. The weekly frequency of resistance training was determined by summing the number of times participants wrote in their diary activities such as “resistance training”, “resistance exercises” or “exercises on gym-machines”.

Calculation of PA maintenance

PA maintenance was calculated as level at 12-month follow-up minus level at the end of the exercise intervention. This calculation was performed separately for moderate intensity aerobic PA, vigorous intensity aerobic PA and moderate-to-vigorous intensity aerobic PA (expressed in minutes/week) as well as for resistance training (expressed in number of sessions/week). The PA maintenance variable was dichotomised and participants were categorised as maintainers (i.e. change in any aerobic PA intensity level ≥ 0 and/or change in resistance training ≥0) or non-maintainers (i.e. maintained neither aerobic PA levels nor resistance training).

Assessment of predictors

The choice and number of variables potentially associated with PA maintenance was based on previous research [18, 21, 25], and restricted by the sample size (i.e. at least five participants per predictor variable in the smallest group) [36]. A total of 12 variables were collected at baseline (prior to randomisation) and 11 at the end of the exercise intervention (6 months after randomisation). Variables were assessed with established and validated methods, and one study-specific item (see below).

Variables collected at baseline

Demographic and health behaviour information was collected by self-report and consisted of age, living situation (Living with a partner vs. Living alone) and tobacco use (Never smoked or used snus vs. Former or current smoker/snus user). Patient-reported variables were collected by self-reports and included anxiety assessed by the subscale of Hospital Anxiety and Depression Scale (HADS-A, range 0–21) [37], and cancer-related fatigue assessed by the subscale general fatigue of Multidimensional Fatigue Inventory (MFI-20, range 4–20) [38]. In the present study, the reliability of these two subscales was confirmed by Cronbach’s alpha values ranging from 0.86 to 0.87. Health-related quality-of-life (HRQoL) was assessed by the subscale of European Organisation for Research and Treatment of Cancer quality of life questionnaire (EORTC QLQ-C30, range 0–100) [39], and exercise self-efficacy by the 9-item Exercise Barrier Self-Efficacy Scale (EBSS, range 0–10) [40]. In the present study, the reliability of these scales was confirmed by Cronbach’s alpha values ranging from 0.85 to 0.87 for the subscale of EORTC QLQ-C30 and 0.90 to 0.91 for exercise self-efficacy. Exercise expectations were assessed with one study-specific item, where participants rated on a 0–10 numerical rating scale how confident they were that exercise could reduce the risk of cancer recurrence (0=‘Not at all confident’ and 10 = ‘Very confident’). We adapted this item from a previously validated template (weighted Kappa coefficient of 0.62) for adults with rheumatoid arthritis [41, 42]. Exercise motivation was assessed with a study-specific questionnaire including three items [43], where participants rated on a 100 mm visual analogue scale 1) how important it was to them 2) how confident, and 3) how ready they were to perform aerobic PA at moderate intensity (0=‘Not at all important/confident /ready’ and 100 = ‘Very important/confident/ready”). The three items were averaged to calculate an overall score. In the present study, the reliability of the overall score was confirmed by Cronbach’s alpha values ranging from 0.83 to 0.90. Objectively assessed variables consisted of cardiorespiratory fitness measured as maximal oxygen uptake (VO2max, mL/kg/min) during walking/running on a treadmill to exhaustion using a modified Balke-protocol [44], and body mass index (BMI, kg/m2) calculated by the study staff the day the VO2max test was performed. Data about (neo-)adjuvant cancer treatment (Chemotherapy vs. No chemotherapy) were extracted from medical records.

Variables collected post-exercise intervention

Patient-reported variables consisted of anxiety, cancer-related fatigue, HRQoL, exercise self-efficacy and exercise motivation, and were assessed as described above. Objectively assessed variables consisted of VO2max, BMI and muscle strength. VO2max and BMI were assessed as described above. Upper and lower muscle strength (kg) were assessed as one repetition maximum including chest press and seated leg press, respectively. Change scores for VO2max, BMI and muscle strength were used for the analyses and were computed by subtracting the baseline score from the post-exercise intervention score. Exercise adherence to the intervention was assessed using training logs filled out by the participants after each session and was calculated for resistance and endurance training respectively, as performed exercise volume divided by prescribed exercise volume (proportion, %) [15].

Statistical analyses

Baseline characteristics across intervention groups were compared using one-way ANOVA or Kruskal-Wallis test for continuous variables and Chi2 test for nominal variables. Difference in baseline characteristics between those included in the analysis and those who were lost to follow-up were examined using independent t-test or Mann-Whitney test for continuous variables and Chi2 test for nominal variables.

PA maintenance at 12-month follow-up is reported using frequencies and proportions. The main effect of self-regulatory BCTs on PA maintenance was examined using multiple logistic regression. The analysis was adjusted for exercise intensity, interaction (exercise intensity x self-regulatory BCTs), study site and cancer diagnosis. All the variables were included in the model using effect coding [45], except for study site and cancer diagnosis that were included using dummy coding. The results are presented as odds ratio (OR) with 95% Confidence Intervals (95% CI).

Predictors of PA maintenance were examined using multiple logistic regression. Two models were run separately; one including baseline variables (Model 1) and one including post-exercise intervention variables (Model 2). Each model also included the intervention groups to adjust for potential differences between the groups. For each model, the variables were entered simultaneous [46]. Dichotomous variables were generated for nominal variables with two categories (i.e. living situation, tobacco use and cancer treatment). Age, anxiety, cancer-related fatigue, HRQoL, exercise self-efficacy, exercise expectations, exercise motivation, VO2max, BMI, muscle strength and exercise adherence were treated as continuous variables with ORs reported per whole unit of the measure (e.g. 1 year, 1 kg/m2 etc). Missing data were ≤ 7% in all variables, except for tobacco use (9.6% missing), VO2max change (10.6% missing) and muscle strength change (12.6% missing). Missing data were not associated with the majority of the baseline characteristics (e.g. age, living situation, education, BMI) or the outcome (PA maintenance) (Suppl. Table 1). Thus, participants with missing data were excluded by listwise deletion from the analysis as this statistical technique is considered to produce unbiased estimates of the exposure-outcome association if missing data are unrelated to the outcome [47]. The results are presented for each model separately. ORs with 95% CI are presented for each variable and indicate the individual contribution of each variable after adjusting for all the others included in the same model.

Assumptions of linearity (continuous variables only), non-multicollinearity (all variables) and goodness of fit were assessed for each analysis. All data analyses were performed with the Statistical Package for the Social Sciences (SPSS, v.27).

Results

Participants

Among the 577 participants randomised to one of the four intervention groups, 316 (55%) provided available SWA and diary data at the two measurement points (i.e. end of the exercise intervention and 12-month follow-up). Of those participants, 301 (95%) performed the required minimum level of PA at the end of the exercise intervention and were included in the analyses.

Participants’ characteristics were similar in the four intervention groups (Table 2). At baseline, participants’ mean age was 59 years (SD 12). The majority were diagnosed with breast cancer (78%) and received chemotherapy as primary (neo-)adjuvant treatment (59%).

Table 2 Baseline characteristics of follow-up participants and participants lost to follow-up in the Phys-Can RCT

The 276 participants (48%) lost to follow-up significantly differed from the follow-up population at baseline in being mainly former or current smokers/snus users (p = 0.006), less physically active (p < 0.001), having a lower VO2max (p < 0.001), a lower HRQoL (p = 0.005), and a higher BMI (p < 0.001).

PA maintenance at 12-month follow-up

The proportion of participants who maintained their PA levels at 12-month follow-up is provided for each intervention group in Table 3. In total, 223 participants (74%) maintained their level of PA at 12-month follow-up, either aerobic PA only (68%), resistance training only (2%) or both aerobic PA and resistance training (4%).

Table 3 Physical activity maintenance at 12-month follow-up in the Phys-Can RCT (n = 301)

Effect of self-regulatory BCTs on PA maintenance

The multiple ordinal logistic regression revealed a main effect of self-regulatory BCTs on PA maintenance at 12-month follow-up (OR = 1.80, 95%CI [1.05–3.08], p = 0.033). A significant interaction was observed between exercise intensity and self-regulatory BCTs (OR = 1.77, 95%CI [1.03–3.04], p = 0.037), indicating that the effect of self-regulatory BCTs was larger in participants in the HI group with self-regulatory BCTs compared to those in the LMI group with self-regulatory BCTs.

Baseline predictors of PA maintenance (Model 1)

Baseline predictors of PA maintenance are presented in Table 4. Sixty-three participants were excluded from the analysis due to missing data (Suppl. Table 1). Participants reporting higher HRQoL (OR = 1.03, 95%CI [1.00–1.06], p = 0.018) or higher exercise motivation (OR = 1.02, 95%CI [1.00–1.04], p = 0.011) at baseline were more likely to maintain their PA levels at 12-month follow-up. In contrast, participants reporting higher expectations to reduce the risk of cancer recurrence with exercise (OR = 0.88, 95%CI [0.78–0.99], p = 0.040) or being former or current smokers/snus users (OR = 0.43, 95%CI [0.21–0.86], p = 0.018) at baseline were less likely to maintain their PA levels at 12-month follow-up.

Table 4 Model 1- multiple ordinal logistic regression model estimating baseline predictors of physical activity maintenancea (n = 238)

Post-exercise intervention predictors of PA maintenance (Model 2)

Post-exercise intervention predictors of PA maintenance are presented in Table 5. Ninety-six participants were excluded from the analysis due to missing data (Suppl. Table 1). Participants who had greater BMI increases over the course of the exercise intervention (OR = 0.63, 95%CI [0.44–0.90], p = 0.010) were less likely to maintain their PA levels at 12-month follow-up.

Table 5 Model 2- multiple ordinal logistic regression model estimating post-exercise intervention predictors of physical activity maintenancea (n = 205)

Discussion

To the best of our knowledge, this is the first study to examine both the effect of self-regulatory BCTs and predictors of long-term PA maintenance in cancer survivors after participating in an exercise intervention. We found that self-regulatory BCTs improved PA maintenance at 12-month follow-up in our study sample, especially in participants exercising at high intensity during the intervention. Further, higher HRQoL and higher exercise motivation at baseline were positively associated with PA maintenance at 12-month follow-up. In contrast, higher exercise expectations, being former or current smokers/snus users at baseline and gaining weight during the exercise intervention were negatively correlated with PA maintenance at 12-month follow-up.

Twelve months after the Phys-Can RCT, the odds of maintaining PA among participants receiving self-regulatory BCTs was 1.8 times the odds of maintaining PA among those who did not receive self-regulatory BCTs. This demonstrates the value of behavioural support to help cancer survivors in maintaining their PA levels in the long term after participating in an exercise intervention. Our results are in line with findings from two recent systematic reviews [8, 13], reporting that self-regulatory BCTs are effective in improving PA maintenance in cancer survivors [8] and in overweight and obese adults [13]. Our findings are also supported by a systematic review involving healthy and clinical adult populations [11], where the use of follow-up prompts (i.e. brief post-intervention contacts to reinforce previous intervention content) was a key strategy to achieve PA maintenance. Thus, it appears that patients are more likely to maintain their PA behaviour if their self-regulatory skills are developed or reinforced during the behaviour change process, which can be achieved using for example goal-setting, review of the behaviour and self-monitoring. Notably, the results from the present study differ from our previous findings, where no effect of the self-regulatory BCTs on exercise adherence was found during the exercise intervention [16]. This lack of effect could be explained by the intervention groups (with and without self-regulatory BCTs) being too similar during the exercise intervention as all were provided with support such as supervised training, social support and feedback. Thus, self-regulatory BCTs as additional support may not have been sufficient to make a difference during the exercise intervention. Another possible explanation is that the self-regulatory BCTs used in the Phys-Can RCT may be more effective in maintaining PA rather than adopting the behaviour [13]. Finally, the large number of lost to follow-up in the present study could also explain the different results, by introducing a selection bias (e.g. participants at follow-up could be those who were more motivated and receptive to this behavioural support). However, it is important to note that our study design does not allow us to determine which self-regulatory BCTs were the most effective in improving long-term PA maintenance in cancer survivors, highlighting the need for further research in broader clinical populations. Furthermore, the majority of our participants did not maintain their level of resistance training at follow-up. It reflects the challenges faced by cancer survivors to continue with this type of PA on their own following a supervised exercise programme, irrespective of the behavioural support provided. Similarly, results from previous studies show a decrease of PA levels following the completion of structured exercise interventions that offered supervised exercise [18, 21, 22, 24]. Besides treatment side-effects, barriers to exercise such as limited access to gym facilities, costs and lack of time are often reported in cancer survivors [48] and could explain the low maintenance levels post-intervention in our study. Therefore, providing participants at the end of exercise interventions with a more individually tailored prescription of resistance training, based on their preferences and needs, could be an adequate alternative to improve long-term PA maintenance of resistance training. For example, exercise barriers such as limited access to gym facilities, costs and lack of time could be overcome with a home-based resistance training programme. The choice of exercise type (e.g. focus on specific muscle groups or more functional exercises) and intensity (e.g. use of weights/resistance bands or merely body weight) should also be discussed with each participant based on their preferences and current capacities. The same principles should be applied regarding exercise frequency and duration with a stepwise progression of exercise.

Regarding predictors, our findings on HRQoL and exercise motivation are in line with the few previous studies investigating long-term PA maintenance in cancer populations. These studies found higher HRQoL [20] and higher exercise motivation [21] to predict PA maintenance in breast cancer survivors at 12-month, 24-month [21], and 5-year [20] follow-up. However, we found only weak associations between HRQoL /exercise motivation and maintenance compared to previous studies, and only regarding the variables at baseline. This could be explained by the high levels and small variation in HRQoL and exercise motivation in our study sample, limiting the possibility to examine the predictive value of these variables on PA maintenance. Notably, exercise self-efficacy did not predict long-term PA maintenance in our sample, which differs from previous studies [10, 24], and could also be explained by the small variation in exercise self-efficacy observed in our study sample. Another explanation is that most of the studies have examined the association between exercise self-efficacy and short-term PA maintenance (less than 12 months) [10, 24] in contrast to our study that focused on long-term PA maintenance. Indeed, findings from previous research indicate that predictors of PA behaviour may differ across the different phases of the behaviour change process (e.g. PA initiation, adoption, short/long-term maintenance) [49]. Interestingly, we found a negative association between baseline exercise expectations and long-term PA maintenance. It is possible that some participants may have had unrealistic expectations to be cured from the disease or feel healthy soon by exercising regularly but were still on treatment or still suffering from treatment-related symptoms at 12-month follow-up. This could have undermined their motivation to continue to be physically active at follow-up. This phenomenon is referred to as the false hope syndrome [50] and implies that high initial exercise expectations may undermine PA maintenance if the expected benefits are small or not attained [50, 51]. However, it is important to note that the item used in this study to measure exercise expectations has been adjusted from a previously validated template, although not validated in cancer survivors [41]. This could partly have impacted on our results and limit the interpretation of our findings. Consequently, given the inconsistent findings across the literature [52], further research is needed to investigate this association. Tobacco use was one of our strongest baseline predictors. This is in contrast with findings from previous studies that have reported no association [21] or borderline associations between smoking status and PA maintenance in cancer survivors [22, 53]. Those divergent results could be explained by the way this variable is usually categorised. In our study, current and former smokers were included in the same category, while in other studies, the variable is often dichotomised as current smokers/non-smokers, which limits the exploration of history of tobacco use as a predictor. In previous studies involving non-cancer populations [54,55,56], smokers and former smokers were found to have poorer health habits than non-smokers, which is partly consistent with our results. Change in BMI during the six-month exercise intervention was another strong predictor, indicating that participants who gained weight during their treatment were less likely to maintain the PA levels at follow-up. Weight gain is a common problem for many cancer survivors, especially for breast cancer survivors treated with adjuvant chemotherapy [57], and several studies have reported that higher weight/BMI is associated with lower exercise levels in cancer survivors [25]. It has been suggested that weight gain can render PA more challenging and less enjoyable to perform [53, 58]. It is also instructive to know that variables that were baseline predictors of PA maintenance in our study, were not post-exercise intervention predictors. This is, for example, the case regarding exercise motivation and could be explained by the measurement method (i.e. we asked our participants about motivation to perform exercise rather than motivation to maintain PA levels) and our definition of maintenance.

Our results may contribute to the development of more effective exercise interventions designed to promote long-term PA behaviour, and guide health professionals working within oncology healthcare. Indeed, this study provides valuable information, describing concrete tools (i.e. specific self-regulatory BCTs) to facilitate long-term PA maintenance in cancer survivors, and identifying patient groups who may benefit from such support. The strengths of the present study include the use of data from a large multicentre RCT, the long-term follow-up and the combination of objective and self-report PA data to capture different aspects of PA [30]. The present study also has limitations, including the large number of participants lost to follow-up and the homogeneous study sample (i.e. mainly women with breast cancer, highly educated, physically active and motivated to exercise), limiting the generalisability of our results. Further, the small number of participants who performed resistance training at 12-month follow-up did not allow us to analyse predictors of aerobic PA and resistance training separately. The outcome (PA maintenance) was therefore dichotomized (maintainers vs. non-maintainers), regardless of the type of PA maintenance (aerobic or resistance). This dichotomisation was deemed adequate as current guidelines stipulate that cancer survivors should perform either aerobic PA or resistance training to achieve similar health benefits [1]. Finally, only participants meeting the exercise guidelines at the end of the exercise intervention were included in the analyses. Using these PA cut-off values has limitations as participants who may have increased their PA levels after the exercise intervention were excluded. However, these criteria were applied to limit selection bias and exclude participants who could otherwise have been categorized as maintainers at 12-month follow-up despite low PA levels at the end of the exercise intervention. It is also important to keep in mind that 95% of participants with available PA data were included in the analyses, indicating a limited impact of applying such PA cut-off values on our results.

Conclusion

Self-regulatory BCTs improved PA maintenance at 12-month follow-up in our study participants. The use of such behavioural support can therefore be recommended to cancer survivors for long-term PA maintenance. Patients with low HRQoL, low exercise motivation, high exercise expectations or with an history of tobacco use at the start of their cancer treatment, as well as those gaining weight during their treatment may be those most in need of such support to maintain PA. However, it remains unclear if similar results would be obtained in more heterogeneous populations and which self-regulatory BCTs are the most effective in improving this complex behaviour. More studies with experimental designs with long-term follow are therefore needed to investigate the efficacy of individual or combinations of BCTs in broader cancer populations.

Availability of data and materials

The data that support the findings of this study are available upon reasonable request.

Abbreviations

BCT:

Behaviour change technique

BMI:

Body mass index

CI:

Confidence interval

EBSS:

Exercise Barrier Self-Efficacy Scale

EORTC QLQ-C30:

European Organisation for Research and Treatment of Cancer quality of life questionnaire C-30

HADS-A:

Hospital Anxiety and Depression scale (Anxiety subscale)

HI:

High intensity exercise

HRQoL:

Health-related quality of life

HRR:

Heart rate reserve

LMI:

Low-to-moderate intensity exercise

MET:

Metabolic Equivalent

MFI:

Multidimensional Fatigue Inventory

OR:

Odds ratio

PA:

Physical activity

Phys-Can:

Physical Training and Cancer

RCT:

Randomised controlled trial

RM:

Repetition maximum

SWA:

SenseWear Armband mini

VO2max:

Maximal oxygen uptake

References

  1. 1.

    Campbell KL, Winters-Stone KM, Wiskemann J, May AM, Schwartz AL, Courneya KS, et al. Exercise guidelines for Cancer survivors: consensus statement from international multidisciplinary roundtable. Med Sci Sports Exerc. 2019;51(11):2375–90.

    PubMed  PubMed Central  Google Scholar 

  2. 2.

    Cormie P, Zopf EM, Zhang X, Schmitz KH. The impact of exercise on cancer mortality, recurrence, and treatment-related adverse effects. Epidemiol Rev. 2017;39(1):71–92.

    PubMed  Google Scholar 

  3. 3.

    Ibrahim EM, Al-Homaidh A. Physical activity and survival after breast cancer diagnosis: meta-analysis of published studies. Med Oncol. 2011;28(3):753–65.

    PubMed  Google Scholar 

  4. 4.

    McTiernan A, Friedenreich CM, Katzmarzyk PT, Powell KE, Macko R, Buchner D, et al. Physical activity in cancer prevention and survival: a systematic review. Med Sci Sports Exerc. 2019;51(6):1252–61.

    PubMed  PubMed Central  Google Scholar 

  5. 5.

    Littman AJ, Tang MT, Rossing MA. Longitudinal study of recreational physical activity in breast cancer survivors. J Cancer Surviv. 2010;4(2):119–27.

    PubMed  Google Scholar 

  6. 6.

    Fassier P, Zelek L, Partula V, Srour B, Bachmann P, Touillaud M, et al. Variations of physical activity and sedentary behavior between before and after cancer diagnosis: results from the prospective population-based NutriNet-Sante cohort. Medicine (Baltimore). 2016;95(40):e4629.

    Google Scholar 

  7. 7.

    Turner RR, Steed L, Quirk H, Greasley RU, Saxton JM, Taylor SJ, et al. Interventions for promoting habitual exercise in people living with and beyond cancer. Cochrane Database Syst Rev. 2018;9:Cd010192.

    PubMed  Google Scholar 

  8. 8.

    Grimmett C, Corbett T, Brunet J, Shepherd J, Pinto BM, May CR, et al. Systematic review and meta-analysis of maintenance of physical activity behaviour change in cancer survivors. Int J Behav Nutr Phys Act. 2019;16(1):37.

    PubMed  PubMed Central  Google Scholar 

  9. 9.

    Vallance JK, Courneya KS, Plotnikoff RC, Dinu I, Mackey JR. Maintenance of physical activity in breast cancer survivors after a randomized trial. Med Sci Sports Exerc. 2008;40(1):173–80.

    PubMed  Google Scholar 

  10. 10.

    Vallance J, Plotnikoff RC, Karvinen KH, Mackey JR, Courneya KS. Understanding physical activity maintenance in breast cancer survivors. Am J Health Behav. 2010;34(2):225–36.

    PubMed  Google Scholar 

  11. 11.

    Fjeldsoe B, Neuhaus M, Winkler E, Eakin E. Systematic review of maintenance of behavior change following physical activity and dietary interventions. Health Psychol. 2011;30(1):99–109.

    PubMed  Google Scholar 

  12. 12.

    Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med. 2013;46(1):81–95.

    PubMed  Google Scholar 

  13. 13.

    Samdal GB, Eide GE, Barth T, Williams G, Meland E. Effective behaviour change techniques for physical activity and healthy eating in overweight and obese adults; systematic review and meta-regression analyses. Int J Behav Nutr Phys Act. 2017;14(1):42.

    PubMed  PubMed Central  Google Scholar 

  14. 14.

    Bandura A. Social cognitive theory of self-regulation. Organ Behav Hum Decis Process. 1991;50:248–87.

    Google Scholar 

  15. 15.

    Demmelmaier I, Brooke HL, Henriksson A, Mazzoni AS, Bjørke ACH, Igelström H, et al. Does exercise intensity matter for fatigue during (neo-)adjuvant cancer treatment? The Phys-can randomized clinical trial. Scand J Med Sci Sports. 2021;31(5):1144–59.

    PubMed  Google Scholar 

  16. 16.

    Mazzoni AS, Brooke HL, Berntsen S, Nordin K, Demmelmaier I. Exercise adherence and effect of self-regulatory behavior change techniques in patients undergoing curative cancer treatment: secondary analysis from the Phys-can randomized controlled trial. Integr Cancer Ther. 2020;19:1534735420946834.

    CAS  PubMed  PubMed Central  Google Scholar 

  17. 17.

    John JM, Haug V, Thiel A. Physical activity behavior from a transdisciplinary biopsychosocial perspective: a scoping review. Sports Med Open. 2020;6(1):49.

    PubMed  PubMed Central  Google Scholar 

  18. 18.

    Schmidt ME, Wiskemann J, Ulrich CM, Schneeweiss A, Steindorf K. Self-reported physical activity behavior of breast cancer survivors during and after adjuvant therapy: 12 months follow-up of two randomized exercise intervention trials. Acta Oncol. 2017;56(4):618–27.

    PubMed  Google Scholar 

  19. 19.

    Lee CE, Von Ah D, Szuck B, Lau YK. Determinants of physical activity maintenance in breast cancer survivors after a community-based intervention. Oncol Nurs Forum. 2016;43(1):93–102.

    PubMed  Google Scholar 

  20. 20.

    Emery CF, Yang HC, Frierson GM, Peterson LJ, Suh S. Determinants of physical activity among women treated for breast cancer in a 5-year longitudinal follow-up investigation. Psychooncology. 2009;18(4):377–86.

    PubMed  PubMed Central  Google Scholar 

  21. 21.

    An KY, Kang DW, Morielli AR, Friedenreich CM, Reid RD, McKenzie DC, et al. Patterns and predictors of exercise behavior during 24 months of follow-up after a supervised exercise program during breast cancer chemotherapy. Int J Behav Nutr Phys Act. 2020;17(1):23.

    PubMed  PubMed Central  Google Scholar 

  22. 22.

    Courneya KS, Friedenreich CM, Reid RD, Gelmon K, Mackey JR, Ladha AB, et al. Predictors of follow-up exercise behavior 6 months after a randomized trial of exercise training during breast cancer chemotherapy. Breast Cancer Res Treat. 2009;114(1):179–87.

    PubMed  Google Scholar 

  23. 23.

    Ungar N, Rupprecht FS, Steindorf K, Wiskemann J, Sieverding M. Worse or even better than expected? Outcome expectancies and behavioral experiences in the context of physical activity among cancer patients. J Health Psychol. 2021;26(5):659–71.

    PubMed  Google Scholar 

  24. 24.

    Loprinzi PD, Cardinal BJ, Si Q, Bennett JA, Winters-Stone KM. Theory-based predictors of follow-up exercise behavior after a supervised exercise intervention in older breast cancer survivors. Support Care Cancer. 2012;20(10):2511–21.

    PubMed  PubMed Central  Google Scholar 

  25. 25.

    Kampshoff CS, Jansen F, van Mechelen W, May AM, Brug J, Chinapaw MJ, et al. Determinants of exercise adherence and maintenance among cancer survivors: a systematic review. Int J Behav Nutr Phys Act. 2014;11:80.

    PubMed  PubMed Central  Google Scholar 

  26. 26.

    Berntsen S, Aaronson NK, Buffart L, Borjeson S, Demmelmaier I, Hellbom M, et al. Design of a randomized controlled trial of physical training and cancer (Phys-can) - the impact of exercise intensity on cancer related fatigue, quality of life and disease outcome. BMC Cancer. 2017;17(1):218.

    PubMed  PubMed Central  Google Scholar 

  27. 27.

    Schmitz KH, Courneya KS, Matthews C, Demark-Wahnefried W, Galvao DA, Pinto BM, et al. American College of Sports Medicine roundtable on exercise guidelines for cancer survivors. Med Sci Sports Exerc. 2010;42(7):1409–26.

    PubMed  PubMed Central  Google Scholar 

  28. 28.

    SenseWear Body Media. BodyMedia SenseWear User Manual Version 8. Pittsburgh: BodyMedia Inc. Available from: https://www.manualslib.com/manual/895732/Bodymedia-Sensewear.html. Cited 29 July 2020.

  29. 29.

    Cereda E, Turrini M, Ciapanna D, Marbello L, Pietrobelli A, Corradi E. Assessing energy expenditure in cancer patients: a pilot validation of a new wearable device. JPEN J Parenter Enteral Nutr. 2007;31(6):502–7.

    PubMed  Google Scholar 

  30. 30.

    Mazzoni AS, Nordin K, Berntsen S, Demmelmaier I, Igelstrom H. Comparison between logbook-reported and objectively-assessed physical activity and sedentary time in breast cancer patients: an agreement study. BMC Sports Sci Med Rehabil. 2017;9:8.

    PubMed  PubMed Central  Google Scholar 

  31. 31.

    U.S. Department of Health and Human Services. Physical Activity Guidelines for Americans. 2nd ed. Washington, DC: Department of Health and Human Services; 2018. Available from: https://health.gov/sites/default/files/2019-09/Physical_Activity_Guidelines_2nd_edition.pdf. Cited 29 December 2020

    Google Scholar 

  32. 32.

    Trost SG, McIver KL, Pate RR. Conducting accelerometer-based activity assessments in field-based research. Med Sci Sports Exerc. 2005;37(11 Suppl):S531–43.

    PubMed  Google Scholar 

  33. 33.

    Herrmann SD, Barreira TV, Kang M, Ainsworth BE. Impact of accelerometer wear time on physical activity data: a NHANES semisimulation data approach. Br J Sports Med. 2014;48(3):278–82.

    PubMed  Google Scholar 

  34. 34.

    Migueles JH, Cadenas-Sanchez C, Ekelund U, Delisle Nyström C, Mora-Gonzalez J, Löf M, et al. Accelerometer data collection and processing criteria to assess physical activity and other outcomes: a systematic review and practical considerations. Sports Med. 2017;47(9):1821–45.

    PubMed  PubMed Central  Google Scholar 

  35. 35.

    Henriksen HB, Berntsen S, Paur I, Zucknick M, Skjetne AJ, Bøhn SK, et al. Validation of two short questionnaires assessing physical activity in colorectal cancer patients. BMC Sports Sci Med Rehabil. 2018;10:8.

    PubMed  PubMed Central  Google Scholar 

  36. 36.

    Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and cox regression. Am J Epidemiol. 2007;165(6):710–8.

    PubMed  Google Scholar 

  37. 37.

    Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the hospital anxiety and depression scale. An updated literature review. J Psychosom Res. 2002;52(2):69–77.

    PubMed  Google Scholar 

  38. 38.

    Hagelin CL, Wengstrom Y, Runesdotter S, Furst CJ. The psychometric properties of the Swedish multidimensional fatigue inventory MFI-20 in four different populations. Acta Oncol. 2007;46(1):97–104.

    PubMed  Google Scholar 

  39. 39.

    Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, 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(5):365–76.

    CAS  Google Scholar 

  40. 40.

    Rogers LQ, Courneya KS, Verhulst S, Markwell S, Lanzotti V, Shah P. Exercise barrier and task self-efficacy in breast cancer patients during treatment. Support Care Cancer. 2006;14(1):84–90.

    PubMed  Google Scholar 

  41. 41.

    Sjöquist ES, Brodin N, Lampa J, Jensen I, Opava CH. Physical activity coaching of patients with rheumatoid arthritis in everyday practice: a long-term follow-up. Musculoskelet Care. 2011;9(2):75–85.

    Google Scholar 

  42. 42.

    Nordgren B, Fridén C, Demmelmaier I, Bergström G, Opava CH. Long-term health-enhancing physical activity in rheumatoid arthritis--the PARA 2010 study. BMC Public Health. 2012;12:397.

    PubMed  PubMed Central  Google Scholar 

  43. 43.

    Miller WR, Rollnick S. Motivational interviewing: helping people change. 3rd ed. New York: Guilford press; 2012.

    Google Scholar 

  44. 44.

    Edvardsen E, Hansen BH, Holme IM, Dyrstad SM, Anderssen SA. Reference values for cardiorespiratory response and fitness on the treadmill in a 20- to 85-year-old population. Chest. 2013;144(1):241–8.

    PubMed  Google Scholar 

  45. 45.

    Kugler KC, Trail JB, Dziak JJ, Collins LM. Effect coding versus dummy coding in analysis of data from factorial experiments. PA: College of Health and Human Development, The Pennsylvania State University; 2012.

    Google Scholar 

  46. 46.

    Harrell FE. Regression modeling strategies: with applications to linear models, logistic regression, and survival analysis. New York: Springer; 2001.

    Google Scholar 

  47. 47.

    Hughes RA, Heron J, Sterne JAC, Tilling K. Accounting for missing data in statistical analyses: multiple imputation is not always the answer. Int J Epidemiol. 2019;48(4):1294–304.

    PubMed  PubMed Central  Google Scholar 

  48. 48.

    Elshahat S, Treanor C, Donnelly M. Factors influencing physical activity participation among people living with or beyond cancer: a systematic scoping review. Int J Behav Nutr Phys Act. 2021;18(1):50.

    PubMed  PubMed Central  Google Scholar 

  49. 49.

    Sutton S. Stage theories of health behaviour. In: Conner M, Norman P, editors. Predicting health behaviour: research and practice with social cognitive models. London: Open University Press; 2005. p. 223–75.

    Google Scholar 

  50. 50.

    Williams DM, Anderson ES, Winett RA. A review of the outcome expectancy construct in physical activity research. Ann Behav Med. 2005;29(1):70–9.

    PubMed  Google Scholar 

  51. 51.

    Wilcox S, Castro CM, King AC. Outcome expectations and physical activity participation in two samples of older women. J Health Psychol. 2006;11(1):65–77.

    PubMed  Google Scholar 

  52. 52.

    Young MD, Plotnikoff RC, Collins CE, Callister R, Morgan PJ. Social cognitive theory and physical activity: a systematic review and meta-analysis. Obes Rev. 2014;15(12):983–95.

    CAS  PubMed  Google Scholar 

  53. 53.

    Courneya KS, Stevinson C, McNeely ML, Sellar CM, Friedenreich CM, Peddle-McIntyre CJ, et al. Predictors of follow-up exercise behavior 6 months after a randomized trial of supervised exercise training in lymphoma patients. Psychooncology. 2012;21(10):1124–31.

    PubMed  Google Scholar 

  54. 54.

    Jackson SE, Brown J, Ussher M, Shahab L, Steptoe A, Smith L. Combined health risks of cigarette smoking and low levels of physical activity: a prospective cohort study in England with 12-year follow-up. BMJ Open. 2019;9(11):e032852.

    PubMed  PubMed Central  Google Scholar 

  55. 55.

    Smith L, Gardner B, Fisher A, Hamer M. Patterns and correlates of physical activity behaviour over 10 years in older adults: prospective analyses from the English longitudinal study of ageing. BMJ Open. 2015;5(4):e007423.

    PubMed  PubMed Central  Google Scholar 

  56. 56.

    Nduaguba SO, Ford KH, Rascati K. The role of physical activity in the association between smoking status and quality of life. Nicotine Tob Res. 2019;21(8):1065–71.

    PubMed  Google Scholar 

  57. 57.

    Vance V, Mourtzakis M, McCargar L, Hanning R. Weight gain in breast cancer survivors: prevalence, pattern and health consequences. Obes Rev. 2011;12(4):282–94.

    CAS  PubMed  Google Scholar 

  58. 58.

    Milne HM, Wallman KE, Guilfoyle A, Gordon S, Corneya KS. Self-determination theory and physical activity among breast cancer survivors. J Sport Exerc Psychol. 2008;30(1):23–38.

    PubMed  Google Scholar 

Download references

Acknowledgements

The authors would like to thank the study participants for their dedication and time.

Funding

Open access funding provided by Uppsala University. This study was part of the Phys-Can RCT, funded by the Swedish Cancer Society (grant numbers 150841, 160483), the Swedish Research Council (grant number KDB/9514), the Nordic Cancer Union (2015) and The Oncology Department Foundations Research Fund in Uppsala (2016, 2017).

Author information

Affiliations

Authors

Contributions

ASM, HLB, SB, KN and ID made substantial contributions to the conception and design of the study, the acquisition, analysis and interpretation of the data. The first draft of the manuscript was written by ASM and all authors critically revised previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Anne-Sophie Mazzoni.

Ethics declarations

Ethics approval and consent to participate

This study was part of the Phys-Can RCT, approved by the Regional Ethical Review Board in Uppsala (EPN D-number 2014/249). All the participants gave informed written consent before participating and were guaranteed confidentiality.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Suppl Table 1.

Physical activity maintenance at 12-month follow-up and baseline characteristics of participants included vs. excluded from the predictor analyses.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Mazzoni, AS., Brooke, H.L., Berntsen, S. et al. Effect of self-regulatory behaviour change techniques and predictors of physical activity maintenance in cancer survivors: a 12-month follow-up of the Phys-Can RCT. BMC Cancer 21, 1272 (2021). https://doi.org/10.1186/s12885-021-08996-x

Download citation

Keywords

  • Behaviour change
  • Behavioural support
  • Cancer survivors
  • Determinants
  • Exercise
  • Maintenance
  • Oncology
  • Self-regulation