Effect of Physical Activity (PA) Interventions On Health-Related Quality of Life (HRQoL) Post Hematopoietic Stem Cell Transplantation (HSCT): A Systematic Review and Meta-Analysis

Recent advances in HSCT have resulted in improved overall and relapse-free survival and decreased incidence of serious complications. Improved survival has prompted an increased focus on enhancing HRQoL. There is an accumulating body of evidence emphasizing that physical activity (PA) may be one modifiable lifestyle factor associated with decreased negative side effects related to HSCT and improved HRQoL. We sought to summarize the available evidence related to the effects of PA participation and HRQoL in HSCT recipients.

We searched electronic databases including Pubmed/MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials and conference proceedings from American Society of Hematology and American Society of Clinical Oncology to April 1, 2012. No language or publication date limits were used. Studies were eligible for inclusion if they were randomized controlled trials (RCTs) or quasi-experimental trials with a control group comparing exercise interventions with placebo-control or standard of care in adult HSCT recipients. Exercise was defined as any form of PA that was performed on a repeated basis for a fixed time frame over a fixed time interval. Included studies were required to have a HRQoL measure as a primary (1^o) or secondary (2^o) outcome. Change in global HRQoL from pre to post-intervention for intervention and control groups were compared using a DerSimonian and Laird random effects model. Heterogeneity was assessed by calculating the Q- and I² statistics. Meta-regression was performed for the following study characteristics: total number (no.) of study participants; mean age; % female; average body mass index (BMI); mean exercise duration in minutes per session; intervention length in weeks; % adherence; type of transplant (allo- or combination of allo- and auto-transplants); and whether HRQoL was a 1^o or 2^o outcome. Sensitivity analysis assessed the cumulative effect of studies by publication year and no. of participants. Influential analysis was conducted to determine change in pooled effect estimate by removing one study at a time. Publication bias was assessed the Begg's rank correlation and Egger's regression test and plot.

Of the 1606 studies originally identified, 7 studies involving 420 patients who underwent HSCT (of which 208 were assigned to exercise or PA intervention) met all inclusion criteria. One study was excluded from the final analysis as it was identified as an extreme outlier. All 6 remaining studies used the European Organization of Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) measurement tool thus weighted mean differences (WMD) were calculated for ease of interpretation. Participation in an exercise intervention resulted in an increase in global HRQoL with a WMD of 4.72 points (95% CI: −0.04, 9.47; P = 0.05). The proportion of variability in WMD attributed to intra-study heterogeneity was 65% (95% CI: 15%, 85%; P = 0.02) and was due primarily to variation in exercise intervention protocols and study quality. Meta-regression identified that whether HRQoL was a 1^o or 2^o outcome significantly affected the pooled effect estimate (P=0.02). Sensitivity analysis did not reveal any important trends. Influential analysis demonstrated that the study with the most influence when removed was Baumann A, where WMD decreased to 2.86 points (95% CI: −1.25, 8.57). There was no statistical evidence of publication bias by assessment with the Egger's regression test (coefficient for bias 2.35, 95% CI −1.90 to, 6.59, P = 0.20) and Begg's rank correlation test (Z-score 0.75 with continuity correction, P = 0.45).

This analysis summarizes the best available evidence regarding effect of exercise participation on overall self-reported HRQoL in adult HSCT recipients. The pooled estimate, after excluding outliers, showed that exercise participation resulted in a 4.72 point increase in the global HRQoL score on the EORTC QLQ-C30 measure, which was borderline statistically significant (P =0.05). This effect estimate is similar in magnitude to pivotal meta-analyses of exercise interventions in breast cancer survivors. Substantial heterogeneity existed among included trials. Larger RCTs with a strong focus on study quality and long-term effects of exercise participation should be conducted.

View largeDownload slide

View largeDownload slide

Close modal

No relevant conflicts of interest to declare.