Background: Cancer-related fatigue (CRF) is a common consequence of breast cancer (BC). Increased moderate to vigorous physical activity (MVPA) is associated with lower CRF among breast cancer survivors (BCS) and is the first-line treatment for CRF. However, CRF is a common barrier to MVPA among BCS. mHealth MVPA interventions might increase MVPA among BCS with CRF, but existing mHealth interventions are not tailored for BCS with CRF. This study examines whether baseline CRF moderates the effect of Fit2Thrive, a remote MVPA intervention for BCS, on daily MVPA.

Methods: Using Multiphase Optimization Strategy methodology, inactive BCS [n=269; Mage=52.5; SD=9.9)] received a core intervention (Fitbit + Fit2Thrive app) and were randomized to “on” or “off” conditions of five additional components: support calls, deluxe app, online gym, buddy, and text messages. Fit2Thrive was delivered for 12 weeks with a 12-week maintenance period. Baseline CRF was measured using the Patient-Reported Outcomes Measurement Information System (PROMIS) fatigue short form. Scores were coded as binary to indicate clinically elevated fatigue (T-score>55). Daily MVPA was measured via accelerometry at baseline, 12 and 24 weeks. Moderating effects of CRF on time and intervention component were estimated using generalized linear mixed effects models with a gamma distribution and log link function.

Results: The moderating effect of baseline CRF on daily MVPA at weeks 12 and 24 was statistically significant; BCS with CRF demonstrated significantly smaller increases in MVPA at 12 (β=-0.27; p=0.0002) and 24 (β=-0.22; p=0.002) weeks. The moderation effect of baseline CRF was significant for text messages at week 12 (β=0.38; p=0.007) and the deluxe Fit2Thrive app at week 24 (β=-0.50; p=0.0005). For BCS with CRF, increases in daily MVPA were greater at week 12 when text messages were “on” vs. “off” and smaller at week 24 when the deluxe app was “on” vs. “off.” Baseline CRF did not demonstrate a significant moderating effect on any other intervention components.

Conclusions: mHealth MVPA interventions designed for the general BCS population might be less efficacious for those with CRF. Effects of individual intervention components on MVPA may vary as a function of baseline CRF status. Future research should explore how to best tailor mHealth interventions for BCS with CRF to optimize MVPA outcomes among this large, in-need group of BCS.