Objectives Randomised controlled trials (RCTs) can evaluate how well a particular exercise programme reduces cancer treatment-related side effects. Adequate design and reporting of the exercise prescription employed in RCTs is central to interpreting study findings and translating effective interventions into practice. Our previous review on the quality and reporting of exercise prescriptions in RCTs in breast cancer survivors revealed several inadequacies. This review similarly evaluates exercise prescriptions used in RCTs in patients with cancers other than the breast.
Methods The literature was searched for RCTs in persons diagnosed with a cancer other than breast. Data were extracted to evaluate the attention to the principles of exercise training in the study design and the reporting of and adherence to the exercise prescription used for the intervention.
Results Of the 33 studies reviewed, none attended to all of the exercise training principles. Specificity was applied by 89%, progression by 26%, overload by 37%, initial values by 26%, diminishing returns by 9% and reversibility by 3%. Only 2 of 33 studies (6%) reported both the exercise prescription in full and adherence to each individual component of the prescription.
Conclusions Application of the principles of training in exercise RCTs of non-breast cancer survivors was incomplete and inconsistent. Given these observations, interpretation of findings from the reviewed studies should consider potential shortcomings in intervention design. Though the prescribed exercise programme was often described, adherence to the entire prescription was rarely reported providing a less accurate picture of dose–response and challenges in translating programmes to community settings.
- Evidence based reviews
- Exercise rehabilitation
- Intervention efficacy
Statistics from Altmetric.com
An estimated 13.7 million cancer survivors are alive in the USA and this number is projected to grow to 18 million by 2022.1 The population prevalence and projected growth of cancer survivors in other developed countries is similarly high. Survival rates are steadily improving with over 64% of cancer survivors alive 5 or more years past cancer diagnosis and up to 15% living 20 years or longer.2 The majority of cancer survivors experience side effects and symptoms related to their cancer and related treatments that can persist long-term or even manifest after treatment ends. Fatigue, declines in physical function, and losses of bone and muscle mass are some of the more common treatment-related sequelae.3 These changes may contribute to the higher rate of fractures, frailty, cardiovascular disease and diabetes among cancer survivors compared to persons without cancer.4
Exercise has been proposed as a rehabilitation strategy to ameliorate the side effects and symptoms resulting from cancer and treatments. Evidence from randomised controlled trials (RCTs) that evaluate the efficacy of exercise to restore the health and functioning of physiologic systems perturbed by cancer treatment has been accumulating for some time5 and has led to published exercise guidelines for cancer survivors from the American College of Sports Medicine6 ,7 and the American Cancer Society.7 However, these guidelines remain general in nature and the evidence for them stems largely from trials conducted in breast cancer survivors. As trials broaden in scope to examine exercise efficacy in survivors of cancer other than breast, a more comprehensive set of exercise guidelines can be developed.
A critical, yet sometimes overlooked feature of exercise intervention research is the appropriate design, application, reporting and interpretation of the experimental exercise intervention. The basic feature of an exercise intervention is the exercise prescription, which can be likened to a medication prescription that details the appropriate type, dose, frequency and timing of therapy. Every exercise efficacy trial should have a carefully developed and reported exercise prescription that meets the ‘FITT’ formula by outlining the Frequency, Intensity, Time and Type of exercise. Each component of the FITT formula should be evidence-based and appropriate to both the intended population and outcome of interest. Reporting the FITT prescription makes the intervention interpretable and reproducible. Similarly important is to report participants’ adherence to each of the prescribed FITT components. Detailed reporting of the dose of exercise received allows for correct interpretation of the study findings, including the dose–response, and provides information about the tolerability and safety of the original prescription.
An appropriate and quality exercise intervention will carefully match the FITT formula to the study goals through attention to basic exercise training principles. The well-established training principle of specificity, progression, overload, initial values, reversibility and diminishing returns should be applied in the design of exercise interventions (table 1).8 Attention to training principles ensures that the intervention is appropriate for the population, specific to the desired outcome, likely to produce maximal benefit in a specified time period, and benefits attributable to the intervention. Interpreting reports from exercise trials with attention to exercise training principles may also avoid erroneous conclusions about a negative trial outcome that may be more attributable to a poorly designed intervention than a lack of exercise efficacy.
Previously, we have reviewed RCTs of exercise in breast cancer survivors to evaluate how thoroughly the principles of exercise training and FITT components of the exercise prescription are reported in a well-studied group of cancer survivors and point out the implications of these gaps.9 Despite the disproportionately larger number of trials in breast cancer survivors, our review revealed many inconsistencies in the application of exercise training principles and reporting of the FITT prescription. As exercise intervention research expands to focus on survivors of cancer other than the breast, we felt it was timely to conduct a similar qualitative review of exercise interventions for studies in other survivor populations. The purpose of this review is to evaluate exercise interventions in survivors of cancer other than breast cancer for (1) attention to the principles of training, (2) reporting of the FITT prescription and (3) adherence to the prescribed FITT components.
The MEDLINE database was searched to 31 December 2011 using search terms related to cancer (eg, cancer, neoplasm or carcinoma) and exercise (exercise, aerobic exercise, resistance exercise, physical activity and walking). The search was limited to English-language papers published in peer-reviewed journals between 1990 and December 2011. Eligible studies had to meet the following criteria (1) RCT design; (2) inclusion of participants with cancer other than breast and (3) Treatment arms must have included either aerobic and/or resistance exercise plus a control arm. Behavioural interventions without one physiologic outcome and/or alternative exercise interventions such as yoga or Pilates, as well as specific therapeutic interventions were excluded. Two reviewers (KWS and SEN) inspected the title and abstract of each identified study to list eligibility. The two lists were compared and any conflicts were resolved by consensus. Papers deemed eligible were then obtained in full and the study design, protocol and outcomes were extracted and evaluated using defined criteria. Each study was assigned a rating for how well it adhered to each principle of exercise training according to the following: ‘+’=clear application of a training principle; ‘NR’ (not reported)=no application of a training principle; or, ‘?’=unclear or inconsistent application of a training principle (eg, in a mixed intervention, using progression for the resistance component but not the aerobic component). The prescribed training programme for each study was abstracted from the methods section according to the FITT formula (frequency, intensity, time and type) and rated using the same system, that is, +, NR, ?. Participant adherence to the exercise prescription was abstracted from the results section using the same FITT formula. The percentage of studies meeting each criterion was calculated, but no other statistical techniques were used.
The literature search revealed 606 potential articles, of which 33 were deemed eligible for this review (figure 1). Study design, primary and secondary outcomes and other relevant information from eligible studies were abstracted from each article (table 2). Fifteen studies (44%) were aerobic exercise trials (AET),10–24 4 studies (12%) were resistance exercise trials (RET),25–28 12 (36%) were combined aerobic and resistance exercise trials (AET+RET)29–40 and 2 were three-arm trials which randomised participants into aerobic, resistance or control groups.41 ,42 For the three-arm trials, aerobic and resistance arms were separated and evaluated as independent AET and RET interventions, for a total of 35 interventions evaluated. Seventeen studies (52%) were conducted during adjuvant treatment (ie, chemotherapy and/or radiation),10 ,12 ,18 ,19 ,21 ,23 ,24 ,28 ,29 ,31 ,33–35 37–39 ,42 9 studies (27%) were conducted after treatment completion11 ,17 ,22 ,25–27 ,30 ,32 ,40 and 7 studies (21%) were conducted across treatment and recovery.13–16 ,20 ,36 ,41 Interventions ranged from 2 to 30 weeks in duration. Twenty studies (61%) targeted a specific cancer diagnosis (ie, prostate cancer),10 ,12 ,13 ,16 ,21–24 26–28 30–34–37 ,42, while the remaining 13 (39%),11 ,14 ,15 ,17–19 ,20 ,25 ,29 ,38–41 included participants with a variety of cancer types.
Application of the principles of exercise training
The principle of specificity was applied appropriately by 89% of studies; 14 of 17 AET (82%),10–19 ,21 ,24 ,41 ,42 all 6 RET (100%)25–28 ,41 ,42 and 11 of 12 AET+RET (92%) (table 3).29–39 ,41 ,42 It was not reported by two AET (12%),20 ,23 and incorrectly applied by one AET (6%)22 and one AET+RET (8%).40 Overall the principle of progression was applied appropriately by 26% of studies; 4 of 17 AET (24%),11 ,16 ,17 ,42 and 5 of 6 RET (83%).25–28 ,42 Of studies reviewed, 11 AET (65%)10 ,12–15 ,18 ,19 ,22–24 ,41 and 8 AET+RET (67%)30–36 ,40 did not report the use of progression, and this principle was incorrectly applied or unclear in 2 AET (12%),20 ,21 1 RET (17%)41 and 4 AET+RET (33%).29 ,37–39 ,41 The principle of overload was applied appropriately by 37% of studies; 8 of 17 AET (47%),11 ,12 ,15–17 ,21 ,23 ,42 4 of 6 RET (67%)27 ,28 ,41 ,42 and 1 of 12 AET+RET (8%).29 ,42 Overload was not applied by two AET (12%).20 ,41 one RET (17%)25 and six AET+RET (50%).30–32 34–36 The application of overload was unclear in 7 of 17 AET (41%),10 ,13 ,14 ,18 ,19 ,22 ,24 1 of 6 RET (17%)26 and 5 of 12 (42%) AET+RET.33 ,37–40 The principle of initial values was applied appropriately by 26% of studies; 3 of 17 AET (18%),19 ,23 ,41 4 of 6 RET (67%)25–27 ,41 and 2 of 12 AET+RET (17%);32 ,36 whereas, 13 of 17 AET (76%)10–12 14–18 20–22 ,24 ,42 and 10 of 12 AET+RET (83%)29–31 33–35 37–39 ,40 ,42 did not apply this principle. It was unclear as to whether exercise was prescribed based on initial values for one AET (6%).13 By conducting multiple assessments during and/or after an intervention the principles of diminishing returns and reversibility can be evaluated. Of the studies reviewed, only one intervention (AET),31 ,36 ,38 appropriately applied the principle of reversibility by following up after the intervention was withdrawn, while five interventions (14%), including two AET16 ,24 and three AET+RET36 ,38 ,39 were unclear in their assessment of reversibility. The principle of diminishing returns was attended to fully by three interventions (9%), (one AET,41 one RET41 and one AET+RET39) by examining multiple time points across the intervention period or reporting both a follow-up outcome assessment and participation in the intervention exercise during the follow-up period. Three interventions (9%) were unclear in their attendance to diminishing returns, including two AET16 ,24 and one AET+RET.36
Reporting of the FITT prescription components
Reporting of the FITT prescription components (eg, frequency, intensity, time and type of exercise) was abstracted from study methods and results (figure 2A). Prescribed frequency was reported by 91% of studies10 ,11–16–,29 ,31 ,32 ,35–42, was not reported in two studies (6%),33 ,34 and not clearly described in one study (3%).30 A target intensity was prescribed in 23 studies (70%),10–19–25 27–29 ,37 ,39 ,40 ,42 but was not reported at all in six (18%) studies26 ,33–36 ,41 and was incompletely reported in four studies (13%).30–32 ,38 A prescribed duration (time or sets/repetitions) was reported in 28 of 33 studies (85%),10 ,11–29 ,31 ,32 ,37–42 but was not reported in three (6%) studies33–35 and was incompletely reported in two studies (6%).30 ,36 A prescribed type or mode of exercise was reported in 31 of 33 studies (94%),10–29 31–35 37–42 but two studies (6%) did not clearly describe the training mode.30 ,36
Adherence to the FITT prescription
Of the studies reviewed, two (6%)15 ,16 reported complete details about adherence of participants to each component of the exercise prescription. Eight studies (24%)20–23 ,30 ,34 ,35 did not report adherence to any of the FITT components (figure 2B). Frequency of exercise (number or percentage of exercise sessions attended) was reported in 19 of 33 studies (58%),10 ,11 ,15 ,16 ,18 ,19 ,25–29 31–33 ,37 ,38 ,40–42 whereas intensity of the exercise achieved was reported in five studies (15%).15 ,16 ,24 ,25 ,31 The completed average duration of exercise bouts was reported in 7 of 33 studies (21%)13–16 ,19 ,24 ,39 and the specific type of exercise completed was reported in 11 of 33 studies (33%).12 ,13 ,15 ,16 ,18 ,19 ,24 ,25 ,38–40 The average number of FITT components reported was similar across intervention settings with an average of 1.5, 1.3 and 1.5 components reported for interventions that were performed at home,13 ,14 ,19 ,24 ,33 ,34 ,36 ,39–41 in a supervised setting10–12 15–18 20–23 25–29 ,37 ,38 ,42 or both at home and in a supervised setting,31–32 ,35 respectively.
The RCTs examined in this review for cancers other than breast have contributed knowledge about the benefits exercise offers to cancer survivors with a variety of diagnoses and varying treatments. However, these same studies have not paid consistent attention to basic principles of exercise training in designing the most appropriate exercise interventions for specified outcomes. With less than one-third of RCTs attending to more than half of the six training principles (figure 3), conclusions about the efficacy of exercise to improve various outcomes remains limited. Though meta-analyses on exercise efficacy in non-breast cancer populations generally support a positive benefit of exercise during or after cancer treatment these reviews rarely examine the quality of prescribed interventions.5 ,43–46 Thus, it remains possible that any null trials are the result of a poorly designed intervention or poor delivery rather than a lack of efficacy, but also that the most optimal programmes have yet to be developed and tested. Similarly problematic is when trials fail to report one or all components of the FITT prescription and how well each component was adhered to. Only two studies (6%) reported both the prescribed amount and adherence to all four components of the exercise prescription. Since it is extremely rare for participants to attend 100% of prescribed training sessions it is similarly likely that they may not perform 100% of the prescribed intensity and duration of training. Thus, without information on both the prescribed and received dose of exercise, developing appropriately dosed exercise prescriptions for an expected level of benefit (ie, response) remains difficult and implementation of evidence-based programming is limited.
Application of the principles of exercise training
The principle of specificity, which states that exercise must challenge a specific system or muscle in order for training adaptations to occur within that system or muscle (table 1), was the most commonly applied training principle across the reviewed trials. Eighty-nine per cent of trials correctly matched the training mode to meet the needs of the population and/or with the primary outcome measures. Exercise trials in non-BCS were more apt to apply the principle of specificity compared to studies in BCS9—a reflection of a better match between training mode and outcome measures in non-BCS trials. The principle of specificity ensures that the type of exercise is the most appropriate to achieve the desired outcome. A proper application of specificity is exemplified by studies that selected resistance exercise as a training mode for prostate cancer survivors on androgen deprivation therapy,28 ,37 which is known to cause muscle loss and weakness.47 ,48 Another example is evaluating the efficacy of a specific upper body resistance training programme to reverse muscle weakness in head and neck cancer survivors as measured by changes in upper extremity function.26 ,27 On the other hand, a mismatch between the intervention and the population need and/or outcome, may erroneously indicate exercise was not effective in improving the outcome measures. Trials that did not convincingly apply the specificity principle either mismatched outcome measures with the intervention type (ie, no strength outcome with AET+RET)20 ,22 ,40 or failed to provide rationale for the selection training mode.23 Either scenario makes it difficult to determine whether the prescribed exercise type was the most appropriate for addressing the research question.
The principle of progression, which states that exercise must continually provide adequate overload over time in order to produce continued improvements, was considered in only 26% of studies reviewed. RET were much more likely than AET or AET+RET to properly progress the exercise programme over time. In our prior review, a similarly low proportion of AET in BCS properly attended to progression.9 Failing to plan progression into an exercise intervention could result in small and/or non-significant improvements in outcomes, especially over the course of a longer intervention. Failure to report a progression plan could result in an improper application of an exercise programme in a community setting that could create a safety concern from overtraining and/or an inadequate stimulus. Important to include in the progression plan is the rate of progression and in our review, if this element was not described or was unclear, a ‘?’ was assigned. Several studies include rates of progression for AET, such as increasing the duration of aerobic exercise by 2 min/week over 10 weeks11 or RET, such as increasing the amount of weight lifted by 5 lbs when a participant is able to complete 12 repetitions at a given weight.42 Challenges with how to monitor exercise intensity during aerobic exercise might explain why AETs were less likely to report training progression. Monitoring heart rate17 and/or setting the relative workrate (ie, % of peak power output) on aerobic equipment such as a cycle ergometer16 can be used to monitor and adjust participant effort. When the prescribed progression plan is reported, investigators may also report on whether participants were able to progress more quickly or more slowly providing an indicator of exercise tolerance by a given population. Progression plans can be developed by modifying the FITT prescription to increase the frequency, intensity and/or duration of the programme over time.
The principle of overload, which states that exercise must be performed above habitual levels in order to induce training adaptations, was correctly applied in just 37% of reviewed trials. This proportion of trials attending to overload is similar to that in our prior review in BCS where just 31% of trials included overload.9 Similar to the progression principle, failure to ensure that a prescribed training programme produces an adequate challenge (eg, greater than habitual activity levels) may result in an ineffective trial or suboptimal improvements. Overload is best achieved by prescribing exercise relative to an individual's measured level of initial fitness, such as a % of measured VO2max or a one-repetition maximum. Prescribing aerobic exercise intensity based on age-predicted fitness (ie, % of age-predicted HRmax) can also meet objective for overload; however, we considered overload unclear in these cases because these estimates are not based on an individual's measured fitness. RET can ensure overload by prescribing training weight relative to a test of maximum strength; however, one study that used elastic bands41 and another that used a specialised pneumatically braked eccentric leg ergometer25 have no comparable measure for maximum capacity and training load, making overload unclear. AET+RET studies often provided no measure of overload at all, thus these results from trials should be interpreted with consideration that an adequate stimulus may not have been achieved.
The principle of initial values, which states that individuals with lower baseline fitness levels will experience greater improvements in fitness from training compared to more fit persons, was only considered in one-quarter of reported trials. When more fit and/or physically active individuals are not rigorously screened out of trials, the potential for a ceiling effect in outcome measures occurs and thus the true efficacy of an intervention for those most in need, that is, less fit or functionally limited, remain unknown.49 Only eight trials considered the principle of initial values by including only participants who were underactive19 ,23 ,25 ,31 ,36 ,41 and/or who demonstrated poor initial values of the outcome measure (ie, mild fatigue36 and/or symptoms of shoulder dysfunction26 ,27) at baseline. Unfortunately, definitions for baseline activity criteria varied widely across these studies and were often vague, making it difficult to compare across studies or whether criterion were stringent enough to limit interventions only to the most sedentary of participants. Donnelly36 provided a compelling example, though, of initial values by limiting participant eligibility to those reporting at least mild fatigue and engaging in less than 60 min/week of moderate intensity physical activity per week.36
Even less attention was paid to the principles of diminishing returns, stating that training-related improvements decrease across a training programme as participants become more fit, and reversibility stating that training-related improvements are lost when participants stop exercising regularly. Evaluating either of these principles requires the inclusion of multiple assessment periods, which may explain why they may not be included in trials due to added expense and participant burden. Studies that evaluated diminishing returns included an interim assessment period during the intervention41 or a follow-up period where changes after the intervention were interpreted in relation to the participant's activity level over follow-up.39 Only one study explicitly withdrew the intervention and repeated measures at follow-up to evaluate the reversibility of training effects,31 whereas other studies received a ‘?’ when they failed to measure the primary outcome at follow-up and/or did not specifically examine changes in outcomes when training stopped.16 ,24 ,31 ,36 ,38 ,39 For example, Mustian et al39 tracked and reported continued participation in AET+RET for 1 month after a 4-week intervention during cancer radiotherapy and also conducted follow-up measures of primary outcomes; however, changes in outcomes were not evaluated among those who reduced or stopped training. Assessment of both diminishing returns and reversibility of an intervention will assist in determining the expected time course for improvement of outcomes and what level of exercise must be sustained to produce long-term benefits.
Reporting of FITT prescription and adherence
While most trials reported the frequency and type of exercise prescribed, one third did not adequately report the complete FITT prescription. Intensity was the FITT component least often reported, with six studies (18%) not reporting intensity and four studies (12%) providing vague descriptions of intensity, that is, exercise at a ‘moderate’ or ‘comfortable pace’. Clearly understanding the intensity used in an exercise prescription is critical for both interpreting study findings (ie, was sufficient overload achieved and progression planned?) and for informing clinical practice. Practitioners will have great difficulty implementing evidence-based interventions if programme components are incompletely or inadequately described. Furthermore, if exercise programmes are subsequently applied at intensities higher or lower than that achieved in the intervention, patient safety could be compromised or an inadequate stimulus delivered.
As important as reporting the prescribed FITT programme is reporting adherence to each FITT component during the intervention. In our review, only two studies (6%) reported adherence to every FITT component and only 30% of studies reported adherence for at least two or more FITT components. We observed similarly poor reporting of FITT adherence in RCTs in breast cancer survivors, pointing to a consistent gap in the exercise literature. Frequency was the most commonly reported FITT component and is the most easily monitored. However, frequency provides little information about the level of effort and duration of exercise achieved across sessions. While it is possible that some of the studies may have adhered very closely to their FITT prescription, in others participants may have performed less exercise and thus received a lower total dose than prescribed. A programme that can produce meaningful benefits at a lower than prescribed level of effort could be appealing to many cancer survivors, but this cannot be determined without thorough reporting. Reporting the actual frequency, intensity, duration and type of exercise performed by participants, will allow readers to more accurately interpret study outcomes, as well as reproduce the intervention in both the research and community settings.
For cancer survivors of varying cancer types and with varying treatment regimens and side effect profiles, a ‘one fits all’ approach to exercise is unlikely to maximally benefit this diverse population. Rather, a collection of exercise prescriptions, based on sound principles of exercise training and with rigorously evaluated details (ie, FITT components), must be developed to match the health needs of different cancer survivors. Attention to the principles of training increases the potential for exercise to maximally benefit those in greatest need of improvement, while failure to apply these principles could underestimate the potential benefit of exercise for persons with cancer. Attention to specificity can be attained by closely matching the type of intervention with the study goals. Overload and progression can be attended to by basing the intervention on individually measured fitness levels and building in increasing levels of effort over time. The trajectory and sustainability of training adaptations are best understood when the principles of diminishing returns and reversibility studies are applied by incorporating interim and follow-up measures of outcomes in trial designs. Future studies should report FITT prescriptions and adherence to every FITT component to aid in the development of specific exercise guidelines for cancer survivors that can be translated safely and effectively into clinical practice and community settings.
Contributors All authors contributed to the manuscript in the review of articles and extraction of data from reviewed articles, creation of tables and figures and writing and review of text and have approved the final version for submission. KM Winters-Stone is responsible for the overall content of the review.
Funding This work received no specific funding. Dr Winters-Stone is supported in part by grants from the National Institutes of Health 1 R01 CA163474-01 and 1 R21 CA164661-01. Ms Neil was supported by an Urban Poling Activator Scholarship, University of British Columbia.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.