Targeting Exercise Interventions to Patients With Cancer in Need: An Individual Patient Data Meta-Analysis

TY - JOUR

T1 - Targeting Exercise Interventions to Patients With Cancer in Need: An Individual Patient Data Meta-Analysis

AU - Buffart, Laurien M

AU - Sweegers, Maike G

AU - May, Anne M

AU - Chinapaw, Mai J

AU - van Vulpen, Jonna K

AU - Newton, Rob U

AU - Galvão, Daniel A

AU - Aaronson, Neil K

AU - Stuiver, Martijn M

AU - Jacobsen, Paul B

AU - Verdonck-de Leeuw, Irma M

AU - Steindorf, Karen

AU - Irwin, Melinda L

AU - Hayes, Sandi

AU - Griffith, Kathleen A

AU - Lucia, Alejandro

AU - Herrero-Roman, Fernando

AU - Mesters, Ilse

AU - van Weert, Ellen

AU - Knoop, Hans

AU - Goedendorp, Martine M

AU - Mutrie, Nanette

AU - Daley, Amanda J

AU - McConnachie, Alex

AU - Bohus, Martin

AU - Thorsen, Lene

AU - Schulz, Karl-Heinz

AU - Short, Camille E

AU - James, Erica L

AU - Plotnikoff, Ronald C

AU - Arbane, Gill

AU - Schmidt, Martina E

AU - Potthoff, Karin

AU - van Beurden, Marc

AU - Oldenburg, Hester S

AU - Sonke, Gabe S

AU - van Harten, Wim H

AU - Garrod, Rachel

AU - Schmitz, Kathryn H

AU - Winters-Stone, Kerri M

AU - Velthuis, Miranda J

AU - Taaffe, Dennis R

AU - van Mechelen, Willem

AU - José Kersten, Marie

AU - Nollet, Frans

AU - Wenzel, Jennifer

AU - Wiskemann, Joachim

AU - Brug, Johannes

AU - Courneya, Kerry S

PY - 2018/11/1

Y1 - 2018/11/1

N2 - Background: Exercise effects in cancer patients often appear modest, possibly because interventions rarely target patients most in need. This study investigated the moderator effects of baseline values on the exercise outcomes of fatigue, aerobic fitness, muscle strength, quality of life (QoL), and self-reported physical function (PF) in cancer patients during and post-treatment.Methods: Individual patient data from 34 randomized exercise trials (n = 4519) were pooled. Linear mixed-effect models were used to study moderator effects of baseline values on exercise intervention outcomes and to determine whether these moderator effects differed by intervention timing (during vs post-treatment). All statistical tests were two-sided.Results: Moderator effects of baseline fatigue and PF were consistent across intervention timing, with greater effects in patients with worse fatigue (Pinteraction = .05) and worse PF (Pinteraction = .003). Moderator effects of baseline aerobic fitness, muscle strength, and QoL differed by intervention timing. During treatment, effects on aerobic fitness were greater for patients with better baseline aerobic fitness (Pinteraction = .002). Post-treatment, effects on upper (Pinteraction < .001) and lower (Pinteraction = .01) body muscle strength and QoL (Pinteraction < .001) were greater in patients with worse baseline values.Conclusion: Although exercise should be encouraged for most cancer patients during and post-treatments, targeting specific subgroups may be especially beneficial and cost effective. For fatigue and PF, interventions during and post-treatment should target patients with high fatigue and low PF. During treatment, patients experience benefit for muscle strength and QoL regardless of baseline values; however, only patients with low baseline values benefit post-treatment. For aerobic fitness, patients with low baseline values do not appear to benefit from exercise during treatment.

AB - Background: Exercise effects in cancer patients often appear modest, possibly because interventions rarely target patients most in need. This study investigated the moderator effects of baseline values on the exercise outcomes of fatigue, aerobic fitness, muscle strength, quality of life (QoL), and self-reported physical function (PF) in cancer patients during and post-treatment.Methods: Individual patient data from 34 randomized exercise trials (n = 4519) were pooled. Linear mixed-effect models were used to study moderator effects of baseline values on exercise intervention outcomes and to determine whether these moderator effects differed by intervention timing (during vs post-treatment). All statistical tests were two-sided.Results: Moderator effects of baseline fatigue and PF were consistent across intervention timing, with greater effects in patients with worse fatigue (Pinteraction = .05) and worse PF (Pinteraction = .003). Moderator effects of baseline aerobic fitness, muscle strength, and QoL differed by intervention timing. During treatment, effects on aerobic fitness were greater for patients with better baseline aerobic fitness (Pinteraction = .002). Post-treatment, effects on upper (Pinteraction < .001) and lower (Pinteraction = .01) body muscle strength and QoL (Pinteraction < .001) were greater in patients with worse baseline values.Conclusion: Although exercise should be encouraged for most cancer patients during and post-treatments, targeting specific subgroups may be especially beneficial and cost effective. For fatigue and PF, interventions during and post-treatment should target patients with high fatigue and low PF. During treatment, patients experience benefit for muscle strength and QoL regardless of baseline values; however, only patients with low baseline values benefit post-treatment. For aerobic fitness, patients with low baseline values do not appear to benefit from exercise during treatment.

KW - Journal Article

U2 - 10.1093/jnci/djy161

DO - 10.1093/jnci/djy161

M3 - SCORING: Journal article

C2 - 30299508

VL - 110

SP - 1190

EP - 1200

JO - JNCI-J NATL CANCER I

JF - JNCI-J NATL CANCER I

SN - 0027-8874

IS - 11

ER -