Exercise and Cancer-Related Fatigue in Adults: Three Randomized Trials for Journal Club

Abstract

This focused review compares three randomized trials of exercise for cancer-related fatigue in adults. The trials span active chemotherapy, neoadjuvant chemotherapy, and gastric-cancer treatment, and therefore address different clinical moments rather than one uniform population. Interventions ranged from a six-week, high-intensity multimodal programme to self-designed metabolic-equivalent exercise over three months and supervised aerobic or resistance training delivered during neoadjuvant chemotherapy. The trials also used different fatigue instruments: the EORTC QLQ-C30, Revised Piper Fatigue Scale, and EORTC QLQ-FA12. Exercise was associated with lower fatigue in the multimodal and gastric-cancer trials, whereas the breast-cancer trial found no clear immediate post-chemotherapy advantage over a delayed-exercise control. The comparison is therefore descriptive rather than a ranking of exercise modalities. The main interpretive problems are heterogeneity of cancer populations and interventions, different fatigue constructs and score directions, incomplete follow-up in two trials, and delayed exercise in the breast-cancer control group. For journal-club discussion, the most defensible conclusion is that exercise may improve fatigue for some adults during or after treatment, but the size and durability of benefit cannot be compared directly across these three studies.

Introduction

Clinical question

Cancer-related fatigue is a patient-reported outcome assessed during active treatment and after treatment in exercise trials. The three trials below were selected because adults were randomised to an exercise-containing intervention and a comparator, and fatigue was measured as an outcome. [citation_1191d8]

Scope of the comparison

The evidence is deliberately narrow: one trial enrolled adults with 21 cancer diagnoses during chemotherapy or treatment for advanced disease, one enrolled adults with gastric cancer, and one enrolled women with breast cancer receiving neoadjuvant chemotherapy. [citation_109h0m] [citation_1su9n1] [citation_1191d8] These populations should not be treated as interchangeable.

Why the measures remain separate

Fatigue was measured with different instruments and at different time points. [citation_109h0m] [citation_1su9n1] [citation_1191d8] The table therefore reports each instrument exactly as used rather than converting scores into a common effect estimate.

1. Multimodal exercise during chemotherapy

Trial population

Adamsen et al. randomised 269 adults with cancer, including 73 men and 196 women, with 21 diagnoses; participants were receiving adjuvant chemotherapy or treatment for advanced disease. The mean age was 47 years, with an age range of 20–65 years. [citation_109h0m]

Exercise and control

The intervention combined supervised high-intensity cardiovascular and resistance training with relaxation, body-awareness training, and massage, delivered for nine hours weekly over six weeks alongside conventional care. The control group received conventional care alone. [citation_109h0m]

Fatigue outcome

Fatigue was the primary outcome within the EORTC QLQ-C30 framework. At six weeks, the adjusted estimated improvement between groups was −6.6 points, with a 95% confidence interval of −12.3 to −0.9 and an effect size of 0.33. [citation_109h0m]

Journal-club limitation

The intervention was not an exercise-only test because it bundled exercise with relaxation, body-awareness training, and massage. In addition, 235 of 269 participants completed follow-up, so the result does not represent the full randomised sample. [citation_109h0m]

2. Metabolic-equivalent exercise in gastric cancer

Trial population

Xin et al. randomised 129 adults with gastric cancer to self-designed metabolic-equivalent exercise or control. The analysed sample was smaller: 119 participants remained after exclusions for medical or personal reasons, loss to follow-up, and intervention discontinuation. [citation_1su9n1]

Exercise and control

The intervention was described as self-designed metabolic-equivalent exercise; the control group received the comparator condition reported by the trial. The retrieved trial report does not provide further control-group detail, so this field should be treated as not reported rather than assumed to mean usual care. [citation_1su9n1]

Fatigue outcome

Fatigue was assessed with the Revised Piper Fatigue Scale, with the EORTC QLQ-C30 also used for quality of life and symptom dimensions. At three months, the total Revised Piper Fatigue Scale score was 2.86 ± 1.75 in the exercise group versus 4.65 ± 1.29 in control, with p = 0.009. [citation_1su9n1]

Journal-club limitation

The three-month result was based on 119 rather than 129 randomised participants, and the retrieved report does not state whether the primary analysis was intention-to-treat. That attrition and the limited control description make the result less secure than a fully reported, retained randomised comparison. [citation_1su9n1]

3. Aerobic or resistance training around neoadjuvant chemotherapy

Trial population

The fatigue analysis of the BENEFIT randomised trial included 184 women with breast cancer scheduled for neoadjuvant chemotherapy; mean age was 50 years with a standard deviation of 11 years. Participants were allocated to aerobic training, resistance training, or a waitlist control. [citation_1191d8]

Exercise and delayed control

The aerobic and resistance groups trained during neoadjuvant chemotherapy with two supervised and one home-based session weekly. The waitlist group completed the same resistance-training programme only after breast surgery. [citation_1191d8]

Fatigue outcome and follow-up

Cancer-related fatigue was measured with the EORTC QLQ-FA12 before chemotherapy, after nine weeks, after chemotherapy and before surgery, six months after surgery, and twelve months after surgery. There was no clear post-intervention difference between either exercise group and the waitlist control; at six months, the waitlist group had more favourable total and physical fatigue values than the aerobic group. [citation_1191d8]

Journal-club limitation

The comparator was delayed exercise rather than continued non-exercise care. That design is useful for comparing timing, but it weakens a claim that exercise during chemotherapy is better than no exercise over the longer follow-up. The fatigue result is also a secondary analysis of the BENEFIT trial. [citation_1191d8]

4. Fatigue measurement across trials

EORTC QLQ-C30

Adamsen et al. used the EORTC QLQ-C30 framework and reported fatigue as the primary outcome. Its six-week estimate should not be numerically compared with the Revised Piper Fatigue Scale score in the gastric-cancer trial. [citation_109h0m]

Revised Piper Fatigue Scale

Xin et al. used the Revised Piper Fatigue Scale and reported a lower total score after three months in the exercise group. The scale and its score distribution should remain separate from the EORTC instruments. [citation_1su9n1]

EORTC QLQ-FA12

The BENEFIT fatigue analysis used the EORTC QLQ-FA12, including total and physical cancer-related fatigue. It assessed repeated time points through twelve months after surgery, unlike the six-week and three-month endpoints reported in the other two trials. [citation_1191d8]

Direction and meaning of scores

The trial reports do not establish one common score direction across all three instruments in the retrieved passages. The comparison therefore uses the authors’ stated improvement, group difference, or between-group conclusion instead of relabelling all scales as if they measured the same numerical construct.

5. Control groups and treatment timing

Conventional care

Adamsen et al. compared the multimodal programme with conventional care alone during chemotherapy or treatment for advanced disease. This gives a clinically familiar comparator, but the bundled intervention prevents attribution of the fatigue result to exercise alone. [citation_109h0m]

Control condition not fully described

Xin et al. report a control group but the retrieved passage does not describe its content in enough detail to classify it confidently as usual care, attention control, or another comparator. The table records this as not reported. [citation_1su9n1]

Waitlist with delayed training

In BENEFIT, the waitlist group was not permanently exercise-naive: it received the same resistance training after surgery. The trial therefore provides a test of exercise timing as well as a short-term comparison with delayed training. [citation_1191d8]

Active treatment versus survivorship

The three trials cover different treatment contexts. Adamsen et al. studied people during chemotherapy or advanced-disease treatment, Xin et al. studied people with gastric cancer, and BENEFIT studied breast cancer during neoadjuvant chemotherapy with post-surgical delayed exercise for controls. [citation_109h0m]

6. What a journal club can reasonably conclude

Direction of findings

Two trials reported lower fatigue after exercise-containing interventions: Adamsen et al. reported an adjusted six-week improvement, and Xin et al. reported a lower three-month Revised Piper Fatigue Scale score. [citation_109h0m] [citation_1su9n1]

Non-confirmatory comparison

The BENEFIT fatigue analysis did not show a clear immediate post-intervention advantage for aerobic or resistance training over the delayed-exercise control, and its later comparison favoured the delayed-training group over aerobic training at six months. [citation_1191d8]

Durability

Only the BENEFIT analysis reported fatigue assessments extending to twelve months after surgery. The Adamsen trial reported six-week follow-up, while the Xin trial reported a three-month assessment; longer-term persistence was not reported in the retrieved passages for either study. [citation_109h0m] [citation_1su9n1] [citation_1191d8]

Bottom line for discussion

The trials support discussing exercise as a plausible fatigue-management strategy, not selecting one superior prescription. The most important questions are whether the intervention is exercise-only, whether the comparator remains inactive, which fatigue construct is measured, and whether benefit persists after treatment ends.

Methodological Comparison

The table keeps the fatigue scales separate and uses “not reported” where the retrieved trial passage does not provide enough detail.

Study	Population	Exercise intervention	Control group	Fatigue measure	Follow-up	One important limitation	Citation
Adamsen et al. (2009), Effect of a multimodal high intensity exercise intervention in cancer patients undergoing chemotherapy	269 adults; 21 cancer diagnoses; 73 men and 196 women; mean age 47 years; adjuvant chemotherapy or treatment for advanced disease	Supervised high-intensity cardiovascular and resistance training plus relaxation, body-awareness training, and massage; 9 hours/week for 6 weeks	Conventional care	EORTC QLQ-C30, with fatigue as the primary outcome	Six weeks; 235/269 completed follow-up	Multicomponent intervention cannot isolate the exercise effect; 34 participants did not complete follow-up	[citation_109h0m]
Xin et al. (2024), The effect of self-designed metabolic equivalent exercises on cancer-related fatigue in patients with gastric cancer	129 adults with gastric cancer randomised; 119 analysed after exclusions and losses	Self-designed metabolic-equivalent exercise; further prescription detail not reported in the retrieved passage	Control group; comparator content not reported in the retrieved passage	Revised Piper Fatigue Scale; EORTC QLQ-C30 also collected	Baseline and 3 months	Post-randomisation exclusions and losses reduced the analysed sample from 129 to 119; intention-to-treat status not reported in the retrieved passage	[citation_1su9n1]
Haussmann et al. (2026), fatigue analysis of the BENEFIT RCT	184 women with breast cancer scheduled for neoadjuvant chemotherapy; mean age 50 years	Aerobic training or resistance training during chemotherapy; two supervised and one home-based session weekly	Waitlist; the same resistance training was provided after breast surgery	EORTC QLQ-FA12, including total and physical cancer-related fatigue	Baseline, 9 weeks, post-chemotherapy/pre-surgery, 6 months after surgery, and 12 months after surgery	Delayed-exercise control is not a permanently inactive comparator; fatigue analysis is secondary	[citation_1191d8]

Research Gaps

Standardised fatigue outcomes

The three trials used different fatigue instruments: EORTC QLQ-C30, Revised Piper Fatigue Scale, and EORTC QLQ-FA12. [citation_109h0m] [citation_1su9n1] [citation_1191d8] Future trials should pre-specify a primary fatigue instrument and report score direction, minimally important differences, and responder analyses.

Better control conditions

A conventional-care control can estimate pragmatic effectiveness, whereas a delayed-exercise control addresses timing. [citation_109h0m] [citation_1191d8] Trials should state whether controls receive attention, education, usual care, or later exercise, because those choices alter the question being answered.

Longer and more complete follow-up

The reviewed trials provide follow-up from six weeks to twelve months, but only the BENEFIT analysis extends clearly beyond the treatment endpoint. [citation_109h0m] [citation_1su9n1] [citation_1191d8] Retention, missing-data handling, and post-intervention exercise should be reported consistently.

Synthesis/Conclusion

Across three adult cancer RCTs, exercise-containing interventions produced mixed but potentially favourable fatigue findings. The evidence is not suitable for ranking exercise types because the populations, interventions, controls, fatigue scales, and follow-up schedules differ. For journal club, the strongest discussion is therefore methodological: what was actually randomised, whether the control remained inactive, which fatigue construct was measured, and whether the analysis preserved the randomised comparison.

Generated Outputs

[SANDBOX_OUTPUT:eabeb36c01247d6f:exercise_cancer_fatigue_journal_club.md]