Cramp, F., & Daniel, J. (2008). Exercise for the management of cancer-related fatigue in adults. Cochrane Database of Systematic Reviews, CD006145. 

Databases searched were the Cochrane Controlled Trials Register (CCTR), MEDLINE, EMBASE, CINAHL, British Nursing Index, AMED, SIGLE, and Dissertation Abstracts International through July 2007.

Twenty-eight published randomized, controlled trials investigating the effect of exercise on cancer-related fatigue in adults were identified.

Outcomes

Fatigue was assessed using various self-report measures, including the Functional Assessment of Cancer Therapy-Fatigue (FACT-F), fatigue subscale of the POMS, the Piper Fatigue Scale, and the Brief Fatigue Inventory (BFI). Only three studies incorporated more than one fatigue outcome measure.

Treatment Evaluated

Mode, intensity, and timing of exercise differed across studies. Thirteen studies investigated home-based/unsupervised exercise programs, whereas 16 studies investigated supervised, institutionally-based exercise programs. Some studies investigating supervised exercise programs encouraged participants to undertake additional home-based exercise. The mode of aerobic exercise included walking, stationary cycling, or a range of modalities. Three studies included strength training as a component of the exercise program, and two studies investigated the outcomes of resistance training in isolation. Two studies included flexibility training as a component of the exercise program, although several studies incorporated routine stretching as part of the warm-up, cool-down, or both. Yoga was investigated in two trials and seated exercise in one. The intensity of exercise varied greatly across studies, with the comparison complicated by differences in the methods (e.g., heart rate monitoring, predicted oxygen uptake, and patient perceived effort) used to monitor intensity in each study. The length of the intervention also varied greatly, ranging from three to 32 weeks; the largest proportion of studies had an exercise intervention duration of 12 weeks.

• There were 2,083 participants. 
• Participants had various cancer diagnoses, although the majority of studies investigated breast cancer only.
• Nineteen of the 28 studies investigated participants with a specific cancer diagnosis and nine investigated participants with a range of diagnoses.
• The time since diagnosis and stage of treatment (active versus posttreatment) varied widely between studies and, in some cases, within studies.

Using posttest means and mean change scores from the pre- posttests, exercise was found to be statistically significantly more effective than the control intervention in the management of fatigue. In separate analyses, dividing studies based on whether the intervention was performed during or following cancer treatment, exercise was consistently statistically significantly found to be more effective for fatigue than the control intervention. Similarly, positive results concerning the effects of exercise for fatigue were also found when only those studies performed in women with breast cancer were examined. Quantitative comparison of fatigue intervention outcomes in other disease-specific groupings was not possible due to limitations in the available data. However, two studies of the effects of exercise on fatigue in prostate cancer populations reported mixed results. A four-week, home-based walking program had no significant effect on fatigue, whereas a 12-week supervised resistance training program produced a statistically significant improvement in the exercise group compared to the control group. No statistically significant improvements in fatigue were observed in patients with colorectal cancer who were prescribed a 16-week, home-based cardiovascular and flexibility program; in patients with lymphoma attending a weekly supervised yoga session; or in patients with multiple myeloma receiving an individualized 32-week, home-based strength and aerobic training program.  Follow-up assessment of long-term outcomes was limited, with 18 of 28 studies failing to assess outcomes beyond the end of the intervention period. From the remaining ten studies that included a follow-up assessment, three did not present the follow-up data in the original publication.

Methodologic quality of the studies was modest (the majority of the studies assigned an Oxford Quality Score of 2 or 3 on the 0–5 scale, with higher scores indicating better methodologic quality).

• All studies lost points in the scoring of methodologic quality due to the failure of the design to conceal group allocation of study participants to the exercise intervention.
• The authors also noted that observer blinding was rarely reported in the included studies.
• Several studies had methodologic flaws, such as providing incomplete descriptions of dropouts or inadequate descriptions of the methods of randomization.
• Although 20 of the 28 studies provided some data on adherence to the exercise interventions, methods for reporting adherence varied between studies, making it difficult to control for adherence when comparing results across studies.