With an estimated 22.2 million new cancer cases world-wide per year by 2030 (1), attention to cancer- and treatment-related symptoms is of great importance. Patients with cancer suffer from a variety of symptoms, of which cancer-related fatigue is one of the most frequently reported and disabling (2). Fatigue is a potential contributing factor to treatment noncompliance, treatment modifications and early discontinuation of treatment, which in turn might have negative impact on clinical outcomes (3). Although fatigue levels are typically highest during active treatment, elevated levels often persist, even up to 5 yr after successful cancer treatment (2,4). With its negative impact on work, daily activities, social activities, and mood, fatigue causes significant impairment in quality of life among patients with cancer (5,6).
Since the late 1980s, exercise has been proposed as a potential intervention for the prevention and reduction of cancer-related fatigue (7). An increasing number of randomized controlled trials (RCT) in patients with cancer and survivors have evaluated the effects of exercise interventions, most of which have included fatigue as one of the main outcomes. Several meta-analyses have confirmed the beneficial effects of exercise interventions on fatigue, both during and after completion of primary cancer treatment (8–12). A recent meta-analysis, comparing effects of pharmaceuticals, exercise, psychological interventions, and combined exercise and psychological interventions, showed that the largest improvement in cancer-related fatigue was achieved by exercise interventions (weighted effect size, 0.30; 95% confidence interval [CI], 0.25–0.36) (9).
With the availability of an effective intervention for diminishing fatigue, an important next step is to investigate: 1) whether or not exercise intervention effects are consistent across subgroups of patients with cancer; and 2) intervention characteristics with largest effect. Previous meta-analyses evaluated the overall exercise intervention effects on fatigue across a wide range of patient groups and interventions (8–12). It is important to determine if there is heterogeneity in responses to exercise interventions by investigating the potential moderating effects of sociodemographic and clinical characteristics. In addition, identifying characteristics of the exercise intervention that maximize the effect of exercise on fatigue will help to target and improve exercise programs. However, most RCT are not adequately powered to identify differences in effects between subgroups with the use of interaction tests. Further, conventional meta-analyses lack the detailed information on individual patient characteristics that is needed for such analyses, resulting in potential ecological bias (i.e., bias that occurs when patient-level interactions are influenced by study-level interactions) (13,14).
Individual patient data meta-analyses offer an opportunity to investigate moderators of intervention effects in a more thorough manner. By merging and synchronizing raw individual patient data from multiple RCT, a large amount of detailed information on patient and intervention characteristics is available, which facilitates testing interactions at the patient-level (13,14). In the current article, we report the results of an analysis of individual patient data from RCT in an effort to identify relevant moderators of the effects of exercise interventions on fatigue levels in patients with cancer.
METHODS
The current study is part of the Predicting OptimaL cAncer RehabIlitation and Supportive care (POLARIS) project (15): an international infrastructure and shared database of RCT investigating exercise and psychosocial intervention effects in patients with cancer (registered in PROSPERO, CRD42013003805). Effects and potential moderators of the effect of exercise on quality of life have been published previously (16). A detailed description of the POLARIS study design, including the method of study identification and selection, has been published (15). The meta-analysis was conducted in accordance with the PRISMA guidelines.
Briefly, principal investigators of 34 exercise RCT worldwide have shared individual patient data. All principal investigators signed a data sharing agreement, stating that they agreed with the POLARIS policy document and were willing to share anonymized data of study participants. All individual studies had received approval from their local ethics committees. Data sets were imported into the POLARIS database and subsequently harmonized according to standardized protocols. Validity checks were performed on improbable or missing values. Details on requested variables, and data and project management can be found in the published study design (15). All exercise RCT in POLARIS that reported fatigue outcomes were included in the current individual patient data meta-analysis.
Quality assessment
Methodological quality for each RCT included in the current analysis was assessed using the “risk-of-bias” tool of the Cochrane Collaboration by two authors independently (MS and LB) (17). The following aspects were graded as high, low or unclear quality: random sequence generation (high quality if random component was used; 25 trials), allocation concealment (high quality if central, computerized allocation or sequentially numbered sealed envelopes were used; 24 trials), incomplete outcome (high quality if intention-to-treat analyses were performed and missing outcome data were <10% or adequate imputation techniques were used; 25 trials), and incomplete reporting (high quality if fatigue was reported such that data could be entered in an aggregate data meta-analysis; 24 trials). In addition, we assessed adherence (high quality if ≥80% of patients attended ≥80% of sessions; 12 trials) and contamination (high quality if no or limited exercise was present in the control group; 7 trials). For full quality assessment of the included trials, we refer to our previous publication (16).
Outcomes
The main outcome of our analysis was fatigue after the completion of the exercise intervention, measured with Multidimensional Fatigue Inventory general fatigue scale (18) (six studies (19–26)), Fatigue Assessment Questionnaire (27) (two studies (28,29)), Schwartz Cancer Fatigue Scale-6 (30) (three studies (31–33)), revised Piper Fatigue Scale (34) (two studies (35,36)), Checklist Individual Strength (37) (one study (38)), Functional Assessment of Cancer Therapy/Functional Assessment of Chronic Illness Therapy—Fatigue scale (39) (eight studies (40–47)), the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Fatigue subscale (48) (five studies (49–53)) and Short Form Health Survey Vitality subscale (54) (four studies (55–58)). If a study used multiple questionnaires to assess fatigue, we used data from 1) the fatigue-specific questionnaire or 2) the fatigue scale of a cancer-specific quality of life questionnaire.
Potential moderators
Potential moderators of the effects of exercise on fatigue were specified based on previous RCT and meta-(regression) analyses (11,24,52,59–61). Potential demographic moderators included age, sex, marital status, and education level (Table 1) (16). Potential clinical moderators included body mass index (BMI), cancer type, treatment type (surgery, chemotherapy, radiotherapy, and hormone therapy), and presence of distant metastases.
;)
TABLE 1: Potential moderators of exercise intervention effects on fatigue.
A selection of intervention characteristics was identified as potential moderators of effective exercise programs (16): timing of intervention in relation to primary cancer treatment, delivery mode, and intervention duration (Table 1). Potential exercise-related moderators included: prescribed exercise frequency, exercise intensity, exercise type, exercise session time (i.e., FITT factors) and exercise volume (i.e., frequency–time) (Table 1). Exercise intensity was scored according to the definitions of the American College of Sports Medicine (ACSM) (62). Exercise volume was dichotomized into <150 min versus ≥150 min·wk−1, corresponding to ACSM’s exercise guidelines for cancer survivors (63).
Statistical analysis
To allow pooling of the different fatigue questionnaires, we calculated z scores for each individual by subtracting the mean score from the individual score at baseline per fatigue questionnaire and dividing the result by the mean standard deviation at baseline per questionnaire. Within our analyses, we used a one-step approach, that is, simultaneously analyzing all observations while accounting for clustering of observations within studies (14). We conducted all analyses according to the intention-to-treat principle. We used linear mixed effects models to analyze the exercise intervention effect on fatigue. The models were adjusted for the baseline value of fatigue and included a random intercept on study level to take clustering of patients within studies into account. The result, a between-group difference in z scores, corresponds to a Cohen’s d effect size. Absolute effects of 0.2 to 0.5 were considered small, 0.5 to 0.8 as moderate and ≥0.8 as large (64).
To examine whether the effect of exercise interventions on fatigue was moderated by patient characteristics, the aforementioned models were extended with interaction terms of the group allocation with demographic and clinical characteristics. To prevent ecological bias for patient-level interactions, we centered the individual values of potential moderators around their mean study values (13). The independent variables in the models were random intercept, group allocation (exercise intervention or control), baseline value of fatigue, centered patient characteristic, and interaction term (centered patient characteristic–group allocation). The potential moderators were examined one-by-one in separate models. We considered a patient-characteristic to be a moderator if the likelihood ratio test (LRT) indicated a statistically significant improvement of the model fit by adding the interaction term.
To identify intervention- and exercise-related moderators, we did not center the characteristics, because these generally do not vary within studies. A similar method as described above was used, with an interaction term between group allocation and the noncentered intervention- or exercise-related characteristic. If this analysis yielded a statistically significant interaction, exercise intervention effects were reported per stratum. For studies with multiple intervention arms with different characteristics, interaction testing for these specific characteristics was not possible. In those situations, we applied a different approach: the main effect models were evaluated using dummy variables for the intervention- or exercise-related characteristics. Again, analyses were performed for each characteristic separately. Because of the statistically significant moderator effect of delivery mode and the differential exercise characteristics between supervised and unsupervised exercise interventions, we investigated exercise-related moderators stratified per delivery mode.
Because the majority of patients were women with breast cancer, we also tested overall effects on fatigue for patients with breast cancer versus other patients (dichotomized). Because no statistically significant interaction was found (P LRT = 0.7) and effect sizes were similar, analyses are presented for all cancer types combined. Statistical significance was set at a probability of P < 0.05 for all analyses. All statistical analyses were performed using IBM SPSS Statistics 21.0 and R version 3.1.1.
RESULTS
Of the 34 exercise RCT included in the POLARIS database, 31 evaluated the effect of exercise interventions on fatigue, representing 4366 individual patients. Of these, 2437 patients were randomized to an exercise intervention group and 1929 to a control group. Baseline demographic and clinical characteristics of patients in the exercise intervention group and control group are presented separately in Table 2. Patients had a mean age of 54.5 (±11.4) yr and a mean BMI of 27.2 (±5.2) kg·m−2. The majority of patients were female (78%) and diagnosed with breast cancer (70%). Only a small proportion of patients (2%) had distant metastases at baseline. Baseline variables were well balanced over the exercise intervention and control groups. Baseline and end of intervention values were available for analysis of fatigue scores for 3846 patients.
TABLE 2: Demographic and clinical characteristics of individual patients from 31 exercise randomized controlled trials included in the meta-analysis.
Included exercise interventions
The studies included in the analysis, published between 2003 and 2017, were carried out in the Netherlands (19–25,38,56), the United States (31–33,36,55,57,58), Australia (40,45,47,49,50,52), Canada (41–44), Germany (26,28,29,51), United Kingdom (35,46), and Norway (53). Sample sizes ranged from 50 to 330 patients. Of the patients allocated to an exercise intervention group, a little over half (50.3%) participated in exercise during primary cancer treatment (Table 3 and Supplemental Digital Content 1, Appendix, https://links.lww.com/MSS/B757). For the majority of patients, the exercise intervention was (partly) supervised (65.3%). Duration of the exercise interventions varied between 8 and 52 wk. Most patients participated in exercise interventions that prescribed exercising twice a week (53.6%) at a moderate-vigorous to vigorous intensity (48.9%), with session duration of 30 to 60 min (51.7%), and an exercise volume of <150 min·wk−1 (63.2%). A combination of aerobic and resistance exercise was the most common exercise type (50.7%). Of the patients allocated to a control group, the majority were assigned to a usual care group (63.9%).
TABLE 3: Intervention and exercise-related characteristics of individual patients from 31 exercise randomized controlled trials included in the meta-analysis.
Effects on fatigue and moderating effects by patient characteristics
Exercise interventions had statistically significant beneficial effects on fatigue (β = −0.17; 95% CI, −0.22 to −0.12]) compared with control (Table 4 and Fig. 1). Our interaction analyses did not reveal statistically significant moderation of demographic or clinical characteristics on the intervention effect on fatigue (Table 4).
TABLE 4: Effects and patient-level moderators of exercise interventions on fatigue.
FIGURE 1: Forest plot of the effects of exercise interventions on fatigue. Exercise effects (between-group differences in z scores) with 95% CI are presented per intervention, in alphabetical order of first author. Unsupervised interventions are presented above the dashed line, and supervised interventions below. Summary estimates for unsupervised interventions, supervised intervention and all interventions are provided.
Intervention- and exercise-related moderators
Timing of the intervention (i.e., during or after primary cancer treatment) was not found to influence the effect of exercise on fatigue (Table 5). Supervised exercise interventions had statistically significantly larger effects on fatigue than unsupervised exercise interventions (difference between subgroups, −0.18; 95% CI, −0.28 to −0.08). Compared with the control group, supervised exercise interventions showed statistically significant improvement of fatigue (β = −0.23; 95% CI, −0.29 to −0.17), while unsupervised exercise interventions did not (β = −0.04; 95% CI −0.13 to 0.04) (Table 5 and Fig. 1). Within the supervised interventions, duration of the exercise intervention was found to moderate the effect of exercise on fatigue. Largest effects on fatigue were observed in the supervised exercise interventions with a duration ≤12 wk (β = −0.29; 95% CI, −0.39 to −0.20), and smallest effects were observed in supervised exercise interventions with duration of >24 wk (β = −0.11; 95% CI, −0.22 to −0.0002). No other intervention- or exercise-related characteristics were identified as moderators of supervised exercise interventions. Within the unsupervised interventions, neither duration of the intervention nor exercise-related characteristics moderated the effect of exercise interventions on fatigue (Table 5).
TABLE 5: Effects and intervention- and exercise-related moderators of exercise on fatigue.
DISCUSSION
The results of our individual patient data meta-analysis of 31 RCT indicate that exercise interventions have statistically significant beneficial effects on fatigue in patients with cancer. For the populations studied, we found no indication that selection of patients based on their demographic or clinical characteristics would lead to different effects of exercise interventions on fatigue. Instead, beneficial effects on fatigue were observed across all subgroups of patients, supporting a role for exercise in clinical practice for patients with cancer. Strongest effects on fatigue were observed in supervised exercise interventions, whereas effects for nonsupervised interventions were nonsignificant.
The beneficial effect of exercise interventions on fatigue that we observed in this study is in line with previous meta-analyses (8–12). The overall effect on fatigue in our study was statistically significant. Clinical relevance, however, was most pronounced for supervised interventions, that showed significant and small effects on fatigue (64). Despite this small clinically relevant effect, the clinical relevance is further underlined by the magnitude of the problem of fatigue, both in terms of the number of patients affected and the negative impact on patients’ lives, combined with the lack of other, more effective treatment options (9).We recently showed that exercise intervention effects on fatigue are larger in patients with worse baseline fatigue levels (65). Exercise intervention effects on fatigue would possibly be larger in trials selecting patients based on their baseline fatigue level. Another explanation for the small effect may be the joint evaluation of different dimensions of fatigue, resulting in a dilution of the effect by the fatigue dimensions that may be less sensitive to exercise (e.g., mental fatigue) (66).
The availability of a large set of individual patient data in our study offered a unique opportunity to investigate if the effect of exercise interventions on fatigue differed significantly across subgroups of patients with cancer. In previous attempts to identify patient-level moderators using meta-regression analyses, chemotherapy and age were found to be statistically significant modifiers of the effect of exercise (59,60). Importantly, in these studies, only published aggregate data (i.e., summary statistics, such as mean age) were available. Although meta-regression techniques can be used to explore moderation of exercise effects, they have several important disadvantages, including a high risk of bias due to the inability to disentangle patient-level heterogeneity from study-level heterogeneity (ecological bias) (13,67). Therefore, the use of individual patient data to investigate the possible influence of patient-level characteristics is considered superior (66). In line with a previous single RCT exploring moderators of exercise effects on fatigue (19), and our individual patient data meta-analysis of moderators of exercise effects on quality of life (16), we did not identify any statistically significant demographic or clinical moderators of the effect of exercise on fatigue in patients with cancer. Accordingly, our findings support the use of exercise interventions for treatment of fatigue across subgroups of patients with cancer.
Of note, the largest group of patients included in the current individual patient data meta-analysis had breast cancer, followed by prostate cancer and hematological malignancies. Although we did not observe a moderating effect of cancer type in our study, the limited number of cancer sites included in the RCT that were part of the present analysis precludes generalizing our results to all cancer types. Also, the large majority of patients were treated with curative intent. Therefore, the effects of exercise on fatigue in the metastatic setting require further investigation.
We observed that supervised exercise interventions had a larger effect on fatigue than unsupervised interventions, which was both statistically significant and clinically relevant. The larger effects of supervised exercise interventions may be explained by psychosocial benefits due to attention and positive feedback on progress in fitness by the physiotherapist or exercise physiologist delivering the intervention. However, RCT comparing a supervised exercise intervention with a supervised relaxation control group, and thus controlling for attention, also showed that effects on fatigue were significantly higher in the exercise group (28,29). An additional feature that may explain the larger effect size of supervised exercise interventions is access to proper equipment, permitting appropriate overload, monitoring and feedback, hence appropriate intensity. As intensity is often higher in supervised programs (e.g., because of safety reasons), teasing out the relative benefit of delivery mode from that of intensity is difficult. In addition, the larger effects of supervised exercise interventions may be explained by better adherence, greater quality in performance of the exercises, selection of different patients and higher fidelity of patient exercise monitoring. Moreover, goals of a supervised exercise program may be different from goals of unsupervised exercise interventions (e.g., increasing physical fitness vs increasing level of daily physical activity).
Our findings suggest that in the patient groups represented in the included RCT, supervised exercise interventions should be preferred over unsupervised interventions in the treatment of fatigue. However, it should be noted that home-based interventions might be preferred by some patients, because they are not able or willing to attend supervised interventions (68). Moreover, unsupervised interventions have been found to exert positive effects on other outcomes, such as physical functioning (16). We recommend that in general, patients with cancer should be prescribed a supervised exercise intervention, particularly prior to, during and the initial three months following cancer treatment when fatigue effects of treatment are greatest. For those who do not have access to a supervised exercise intervention, unsupervised exercise interventions could still be useful, and might be augmented with e-Health applications. Further investigation is needed to understand which components are the most critical for inclusion in home-based interventions (e.g., including more tailored exercise advice).
Incorporation of supervised exercise into standard care might possibly be more demanding in terms of resource allocation and costs. However, a recent trial showed that a home-based, low-intensity physical activity program was not likely to be cost-effective for fatigue in comparison with usual care, whereas a supervised, moderate-to-high intensity exercise program could be considered cost-effective for fatigue depending on the decision-makers’ willingness-to-pay (69). Also, our finding that exercise interventions with a duration as short as 12 wk already have clinically relevant effects on fatigue may further support the feasibility of incorporating supervised exercise into standard care.
The largest effects of supervised interventions on fatigue were observed in the studies with shortest intervention duration. It is possible that adherence to the intervention, and consequently the effect of exercise on fatigue, decreases over time (70). At the same time, contamination (adoption of exercise by the control group) may increase over time as well (71). In addition, there may be a ceiling effect of exercise interventions on fatigue in cancer patients whereby 12 wk or less is sufficient to counteract disease and treatment detriments and further duration provides maintenance. Furthermore, we cannot exclude the possibility that this finding is partly due to the distribution of other exercise-related characteristics over the duration strata. It would be interesting to compare the long-term fatigue outcomes between the interventions with different durations, but as only a few studies have examined maintenance of intervention effects in the long-term (21,72–74), this remains to be investigated. In any case, it is an important finding that interventions with a duration as short as 12 wk already have positive and clinically relevant effects.
Several methodological limitations of our study should be noted. First, our literature search was conducted in 2012 (15), but we also included published study designs in our literature search and contacted the principal investigators of these studies. Therefore, we included 13 studies that were completed after 2012. Although the literature search focused on quality of life as primary or secondary outcome, fatigue was assessed in most of the RCT (31/34) in parallel with quality of life. Our main aim was to assess moderators of exercise effects on fatigue using individual patient data, and we have no reason to believe that adding more recent studies would significantly change our conclusions regarding patient- and disease-related moderators. As technology is continuously evolving, intervention characteristics of more recent studies might differ from studies conducted in the past. Although technology developments likely take place throughout the whole field of exercise interventions, we cannot exclude the possibility that adding more recent data could impact some intervention and exercise-related moderators.
Second, despite the inclusion of a large amount of individual patient data, the statistical power to detect intervention- and exercise-related moderators was limited because these variables are defined at the study-level (14). Especially the identification of significant exercise-related moderators (FITT factors), which were stratified by delivery mode, may have been compromised by limited power or little variation across studies. Also, we only examined single interactions, but there may be more complex multilevel interactions. Furthermore, within meta-analyses, in general, both patient-level and study-level characteristics can be influenced by (other) study-level characteristics. Centering of patient-level characteristics enabled us to reduce the risk of ecological bias by separating patient-level heterogeneity from study-level heterogeneity when evaluating demographic and clinical moderators (13). This approach was not possible for the analyses on intervention- and exercise-related characteristics, making these analyses more prone to influences of other study-level characteristics. Thus, more RCT that include head-to-head comparisons of intervention and exercise-related characteristics are warranted to confirm our findings and to better disentangle the effects of different study-level determinants, for example, comparisons between exercise interventions with different exercise types (75,76) or posttreatment exercise interventions with different timing in relation to cancer treatment.
Third, adherence to exercise interventions was unknown in the majority of included studies and consequently, patients in the exercise groups might have actually been exposed to different exercise-related characteristics than assumed. In addition, information on contamination was limited, hampering our ability to take the activity level of patients in the control group into account. Because adherence and contamination may affect intervention outcomes (41,77), care should be taken to accurately registering both items to optimally interpret outcomes of exercise interventions.
The present study is the first to collect, synchronize, pool, and analyze individual patient data on cancer-related fatigue from exercise RCT worldwide. We applied a careful standardization of the outcome data and uniform analytic procedures across all studies. An important strength of the study is the availability of a large amount of individual patient data, which enabled us to study multiple demographic and clinical patient-level moderators.
CONCLUSIONS
In conclusion, we found that exercise has statistically significant beneficial effects on fatigue in patients with cancer. These benefits are consistent across subgroups formed on the basis of demographic and clinical characteristics. The effect of exercise interventions on fatigue is significantly larger when performed under supervision. Differential effects of duration and potential roles of adherence and contamination in these findings need further exploration. Our results support implementation of exercise, preferably supervised exercise interventions, in clinical practice.
The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation, and results of the present study do not constitute endorsement by ACSM.
Conflicts of interest: None declared.
Source of funding: The POLARIS study is supported by the Bas Mulder Award, granted to L. M. Buffart by the Alpe d’HuZes foundation/Dutch Cancer Society (VU2011–5045). The contribution of J. K. van Vulpen is financially supported by the World Cancer Research Fund The Netherlands (WCRF NL, project number 2013/997).
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the (U.S.) National Institutes of Health.
REFERENCES
1. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study.
Lancet Oncol. 2012;13(8):790–801.
2. Bower JE. Cancer-related fatigue-mechanisms, risk factors, and treatments.
Nat Rev Clin Oncol. 2014;11(10):597–609.
3. Basch E, Reeve BB, Mitchell SA, et al. Development of the National Cancer Institute’s Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (Pro-Ctcae).
J Natl Cancer Inst. 2014;106(9).
4. Abrahams HJ, Gielissen MF, Schmits IC, Verhagen CA, Rovers MM, Knoop H. Risk factors, prevalence, and course of severe fatigue after breast cancer treatment: a meta-analysis involving 12 327 breast cancer.
Ann Oncol. 2016;27(6):965–74.
5. Curt GA, Breitbart W, Cella D, et al. Impact of cancer-related fatigue on the lives of patients: new findings from the fatigue coalition.
Oncologist. 2000;5(5):353–60.
6. Bower JE, Ganz PA, Desmond KA, Rowland JH, Meyerowitz BE, Belin TR. Fatigue in breast cancer survivors: occurrence, correlates, and impact on quality of life.
J Clin Oncol. 2000;18(4):743–53.
7. MacVicar MG, Winningham ML. Promoting the functional capacity of cancer patients.
Cancer Bull. 1986;38:235–9.
8. Hilfiker R, Meichtry A, Eicher M, et al. Exercise and other non-pharmaceutical interventions for cancer-related fatigue in patients during or after cancer treatment: a systematic review incorporating an indirect-comparisons meta-analysis.
Br J Sports Med. 2018;52(10):651–8.
9. Mustian KM, Alfano CM, Heckler C, et al. Comparison of pharmaceutical, psychological, and exercise treatments for cancer-related fatigue: a meta-analysis.
JAMA Oncol. 2017;3(7):961–8.
10. Tian L, Lu HJ, Lin L, Hu Y. Effects of aerobic exercise on cancer-related fatigue: a meta-analysis of randomized controlled trials.
Support Care Cancer. 2016;24(2):969–83.
11. Cramp F, Byron-Daniel J. Exercise for the management of cancer-related fatigue in adults.
Cochrane Database Syst Rev. 2012;11:CD006145.
12. Puetz TW, Herring MP. Differential effects of exercise on cancer-related fatigue during and following treatment: a meta-analysis.
Am J Prev Med. 2012;43(2):e1–24.
13. Debray TP, Moons KG, van Valkenhoef G, et al. Get real in individual participant data (IPD) meta-analysis: a review of the methodology.
Res Synth Methods. 2015;6(4):293–309.
14. Riley RD, Lambert PC, Abo-Zaid G. Meta-analysis of individual participant data: rationale, conduct, and reporting.
BMJ. 2010;340:c221.
15. Buffart LM, Kalter J, Chinapaw MJ, et al. Predicting optimal cancer rehabilitation and supportive care (Polaris): rationale and design for meta-analyses of individual patient data of randomized controlled trials that evaluate the effect of physical activity and psychosocial interventions on health-related quality of life in cancer survivors.
Syst Rev. 2013;2:75.
16. Buffart LM, Kalter J, Sweegers MG, et al. Effects and moderators of exercise on quality of life and physical function in patients with cancer: an individual patient data meta-analysis of 34 RCTs.
Cancer Treat Rev. 2017;52:91–104.
17. Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane collaboration’s tool for assessing risk of bias in randomised trials.
BMJ. 2011;343:d5928.
18. Smets EM, Garssen B, Bonke B, De Haes JC. The multidimensional fatigue inventory (MFI) psychometric qualities of an instrument to assess fatigue.
J Psychosom Res. 1995;39(3):315–25.
19. Kampshoff CS, Chinapaw MJ, Brug J, et al. Randomized controlled trial of the effects of high intensity and low-to-moderate intensity exercise on physical fitness and fatigue in cancer survivors: results of the resistance and endurance exercise after chemotherapy (react) study.
BMC Med. 2015;13:275.
20. Persoon S, Chin AMJM, Buffart LM, et al. Randomized controlled trial on the effects of a supervised high intensity exercise program in patients with a hematologic malignancy treated with autologous stem cell transplantation: results from the exist study.
PLoS One. 2017;12(7):e0181313.
21. Travier N, Velthuis MJ, Steins Bisschop CN, et al. Effects of an 18-week exercise programme started early during breast cancer treatment: a randomised controlled trial.
BMC Med. 2015;13:121.
22. van Vulpen JK, Velthuis MJ, Steins Bisschop CN, et al. Effects of an exercise program in colon cancer patients undergoing chemotherapy.
Med Sci Sports Exerc. 2016;48(5):767–75.
23. van Waart H, Stuiver MM, van Harten WH, et al. Recruitment to and pilot results of the paces randomized trial of physical exercise during adjuvant chemotherapy for colon cancer.
Int J Colorectal Dis. 2018;33(1):29–40.
24. van Waart H, Stuiver MM, van Harten WH, et al. Effect of low-intensity physical activity and moderate- to high-intensity physical exercise during adjuvant chemotherapy on physical fitness, fatigue, and chemotherapy completion rates: results of the paces randomized clinical trial.
J Clin Oncol. 2015;33(17):1918–27.
25. van Weert E, May AM, Korstjens I, et al. Cancer-related fatigue and rehabilitation: a randomized controlled multicenter trial comparing physical training combined with cognitive-behavioral therapy with physical training only and with no intervention.
Phys Ther. 2010;90(10):1413–25.
26. Wiskemann J, Dreger P, Schwerdtfeger R, et al. Effects of a partly self-administered exercise program before, during, and after allogeneic stem cell transplantation.
Blood. 2011;117(9):2604–13.
27. Glaus A, Muller S. Measuring Fatigue of Cancer patients in the German-speaking region: development of the fatigue assessment questionnaire.
Pflege. 2001;14(3):161–70.
28. Schmidt ME, Wiskemann J, Armbrust P, Schneeweiss A, Ulrich CM, Steindorf K. Effects of resistance exercise on fatigue and quality of life in breast cancer patients undergoing adjuvant chemotherapy: a randomized controlled trial.
Int J Cancer. 2015;137(2):471–80.
29. Steindorf K, Schmidt ME, Klassen O, et al. Randomized, controlled trial of resistance training in breast cancer patients receiving adjuvant radiotherapy: results on cancer-related fatigue and quality of life.
Ann Oncol. 2014;25(11):2237–43.
30. Schwartz AL. The Schwartz Cancer fatigue scale: testing reliability and validity.
Oncol Nurs Forum. 1998;25(4):711–7.
31. Winters-Stone KM, Dobek J, Bennett JA, Nail LM, Leo MC, Schwartz A. The effect of resistance training on muscle strength and physical function in older, postmenopausal breast cancer survivors: a randomized controlled trial.
J Cancer Surviv. 2012;6(2):189–99.
32. Winters-Stone KM, Dobek J, Nail LM, et al. Impact + resistance training improves bone health and body composition in prematurely menopausal breast cancer survivors: a randomized controlled trial.
Osteoporos Int. 2013;24(5):1637–46.
33. Winters-Stone KM, Dobek JC, Bennett JA, et al. Resistance training reduces disability in prostate cancer survivors on androgen deprivation therapy: evidence from a randomized controlled trial.
Arch Phys Med Rehabil. 2015;96(1):7–14.
34. Piper BF, Dibble SL, Dodd MJ, Weiss MC, Slaughter RE, Paul SM. The revised Piper Fatigue Scale: psychometric evaluation in women with breast cancer.
Oncol Nurs Forum. 1998;25(4):677–84.
35. Daley AJ, Crank H, Saxton JM, Mutrie N, Coleman R, Roalfe A. Randomized trial of exercise therapy in women treated for breast cancer.
J Clin Oncol. 2007;25(13):1713–21.
36. Griffith K, Wenzel J, Shang J, Thompson C, Stewart K, Mock V. Impact of a walking intervention on cardiorespiratory fitness, self-reported physical function, and pain in patients undergoing treatment for solid tumors.
Cancer. 2009;115(20):4874–84.
37. Vercoulen JH, Swanink CM, Fennis JF, Galama JM, van der Meer JW, Bleijenberg G. Dimensional assessment of chronic fatigue syndrome.
J Psychosom Res. 1994;38(5):383–92.
38. Goedendorp MM, Peters ME, Gielissen MF, et al. Is increasing physical activity necessary to diminish fatigue during cancer treatment? Comparing cognitive behavior therapy and a brief nursing intervention with usual care in a multicenter randomized controlled trial.
Oncologist. 2010;15(10):1122–32.
39. Yellen SB, Cella DF, Webster K, Blendowski C, Kaplan E. Measuring Fatigue and Other anemia-related symptoms with the functional assessment of cancer therapy (FACT) measurement system.
J Pain Symptom Manage. 1997;13(2):63–74.
40. Cormie P, Galvao DA, Spry N, et al. Can supervised exercise prevent treatment toxicity in patients with prostate cancer initiating androgen-deprivation therapy: a randomised controlled trial.
BJU Int. 2015;115(2):256–66.
41. Courneya KS, Friedenreich CM, Quinney HA, Fields AL, Jones LW, Fairey AS. A randomized trial of exercise and quality of life in colorectal cancer survivors.
Eur J Cancer Care (Engl). 2003;12(4):347–57.
42. Courneya KS, Mackey JR, Bell GJ, Jones LW, Field CJ, Fairey AS. Randomized controlled trial of exercise training in postmenopausal breast cancer survivors: cardiopulmonary and quality of life outcomes.
J Clin Oncol. 2003;21(9):1660–8.
43. Courneya KS, Segal RJ, Mackey JR, et al. Effects of aerobic and resistance exercise in breast cancer patients receiving adjuvant chemotherapy: a multicenter randomized controlled trial.
J Clin Oncol. 2007;25(28):4396–404.
44. Courneya KS, Sellar CM, Stevinson C, et al. Randomized controlled trial of the effects of aerobic exercise on physical functioning and quality of life in lymphoma patients.
J Clin Oncol. 2009;27(27):4605–12.
45. Hayes SC, Rye S, Disipio T, et al. Exercise for health: a randomized, controlled trial evaluating the impact of a pragmatic, translational exercise intervention on the quality of life, function and treatment-related side effects following breast cancer.
Breast Cancer Res Treat. 2013;137(1):175–86.
46. Mutrie N, Campbell AM, Whyte F, et al. Benefits of supervised group exercise programme for women being treated for early stage breast cancer: pragmatic randomised controlled trial.
BMJ. 2007;334(7592):517.
47. Short CE, James EL, Girgis A, D’Souza MI, Plotnikoff RC. Main outcomes of the move more for life trial: a randomised controlled trial examining the effects of tailored-print and targeted-print materials for promoting physical activity among post-treatment breast cancer survivors.
Psychooncology. 2015;24(7):771–8.
48. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer Qlq-C30: a quality-of-life instrument for use in international clinical trials in oncology.
J Natl Cancer Inst. 1993;85(5):365–76.
49. Galvao DA, Spry N, Denham J, et al. 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.
Eur Urol. 2014;65(5):856–64.
50. Galvao DA, Taaffe DR, Spry N, Joseph D, Newton RU. Combined resistance and aerobic exercise program reverses muscle loss in men undergoing androgen suppression therapy for prostate cancer without bone metastases: a randomized controlled trial.
J Clin Oncol. 2010;28(2):340–7.
51. Mehnert A, Veers S, Howaldt D, Braumann KM, Koch U, Schulz KH. Effects of a physical exercise rehabilitation group program on anxiety, depression, body image, and health-related quality of life among breast cancer patients.
Onkologie. 2011;34(5):248–53.
52. Taaffe DR, Newton RU, Spry N, et al. Effects of different exercise modalities on fatigue in prostate cancer patients undergoing androgen deprivation therapy: a year-long randomised controlled trial.
Eur Urol. 2017;72(2):293–9.
53. Thorsen L, Skovlund E, Stromme SB, Hornslien K, Dahl AA, Fossa SD. Effectiveness of physical activity on cardiorespiratory fitness and health-related quality of life in young and middle-aged cancer patients shortly after chemotherapy.
J Clin Oncol. 2005;23(10):2378–88.
54. Ware JE Jr, Gandek B. Overview of the Sf-36 Health Survey and the International Quality of Life Assessment (IQOLA) Project.
J Clin Epidemiol. 1998;51(11):903–12.
55. Cadmus LA, Salovey P, Yu H, Chung G, Kasl S, Irwin ML. Exercise and Quality of life during and after treatment for breast cancer: results of two randomized controlled trials.
Psychooncology. 2009;18(4):343–52.
56. Duijts SF, van Beurden M, Oldenburg HS, et al. Efficacy of cognitive behavioral therapy and physical exercise in alleviating treatment-induced menopausal symptoms in patients with breast cancer: results of a randomized, controlled, multicenter trial.
J Clin Oncol. 2012;30(33):4124–33.
57. Irwin ML, Varma K, Alvarez-Reeves M, et al. Randomized controlled trial of aerobic exercise on insulin and insulin-like growth factors in breast cancer survivors: the Yale Exercise and Survivorship Study.
Cancer Epidemiol Biomarkers Prev. 2009;18(1):306–13.
58. Speck RM, Gross CR, Hormes JM, et al. Changes in the body image and relationship scale following a one-year strength training trial for breast cancer survivors with or at risk for lymphedema.
Breast Cancer Res Treat. 2010;121(2):421–30.
59. Brown JC, Huedo-Medina TB, Pescatello LS, Pescatello SM, Ferrer RA, Johnson BT. Efficacy of exercise interventions in modulating cancer-related fatigue among adult cancer survivors: a meta-analysis.
Cancer Epidemiol Biomarkers Prev. 2011;20(1):123–33.
60. Carayol M, Delpierre C, Bernard P, Ninot G. Population-, intervention- and methodology-related characteristics of clinical trials impact exercise efficacy during adjuvant therapy for breast cancer: a meta-regression analysis.
Psychooncology. 2015;24(7):737–47.
61. Velthuis MJ, Agasi-Idenburg SC, Aufdemkampe G, Wittink HM. The effect of physical exercise on cancer-related fatigue during cancer treatment: a meta-analysis of randomised controlled trials.
Clin Oncol (R Coll Radiol). 2010;22(3):208–21.
62. American College of Sports Medicine.
ACSM’s Guidelines for Exercise Testing and Prescription. 9th ed. Philadelphia: Lippincot Williams and Wilkins; 2014.
63. Schmitz KH, Courneya KS, Matthews C, et al. American College of Sports Medicine Roundtable on Exercise Guidelines for Cancer Survivors.
Med Sci Sports Exerc. 2010;42(7):1409–26.
64. Cohen J.
Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Earlbaum; 1988.
65. Buffart LM, Sweegers MG, May AM, et al. Targeting exercise interventions to patients with cancer in need: an individual patient data meta-analysis.
J Natl Cancer Inst. 2018;110(11):1190–200.
66. van Vulpen JK, Peeters PH, Velthuis MJ, van der Wall E, May AM. Effects of physical exercise during adjuvant breast cancer treatment on physical and psychosocial dimensions of cancer-related fatigue: a meta-analysis.
Maturitas. 2016;85:104–11.
67. Berlin JA, Santanna J, Schmid CH, Szczech LA, Feldman HI. Individual patient- versus group-level data meta-regressions for the investigation of treatment effect modifiers: ecological bias rears its ugly head.
Stat Med. 2002;21(3):371–87.
68. Hardcastle SJ, Maxwell-Smith C, Kamarova S, Lamb S, Millar L, Cohen PA. Factors influencing non-participation in an exercise program and attitudes towards physical activity amongst cancer survivors.
Support Care Cancer. 2018;26(4):1289–95.
69. van Waart H, van Dongen JM, van Harten WH, et al. Cost-utility and cost-effectiveness of physical exercise during adjuvant chemotherapy.
Eur J Health Econ. 2018;19(6):893–904.
70. Kirkham AA, Bonsignore A, Bland KA, et al. Exercise prescription and adherence for breast cancer: one size does not fitt all.
Med Sci Sports Exerc. 2018;50(2):177–86.
71. Hertogh EM, Schuit AJ, Peeters PH, Monninkhof EM. Noncompliance in lifestyle intervention studies: the instrumental variable method provides insight into the bias.
J Clin Epidemiol. 2010;63(8):900–6.
72. Witlox L, Hiensch AE, Velthuis MJ, et al. Four-year effects of exercise on fatigue and physical activity in patients with cancer.
BMC Med. 2018;16(1):86.
73. Schmidt ME, Wiskemann J, Schneeweiss A, Potthoff K, Ulrich CM, Steindorf K. Determinants of physical, affective, and cognitive fatigue during breast cancer therapy and 12 months follow-up.
Int J Cancer. 2018;142(6):1148–57.
74. May AM, Korstjens I, van Weert E, et al. Long-term effects on cancer survivors’ quality of life of physical training versus physical training combined with cognitive-behavioral therapy: results from a randomized trial.
Support Care Cancer. 2009;17(6):653–63.
75. Alibhai SM, Santa Mina D, Ritvo P, et al. A Phase II RCT and economic analysis of three exercise delivery methods in men with prostate cancer on androgen deprivation therapy.
BMC Cancer. 2015;15:312.
76. Mijwel S, Backman M, Bolam KA, et al. Adding high-intensity interval training to conventional training modalities: optimizing health-related outcomes during chemotherapy for breast cancer: the Optitrain Randomized Controlled Trial.
Breast Cancer Res Treat. 2018;168(1):79–93.
77. DiMatteo MR, Giordani PJ, Lepper HS, Croghan TW. Patient adherence and medical treatment outcomes: a meta-analysis.
Med Care. 2002;40(9):794–811.
