Effects of the Fit2Thrive Technology-Supported Physical Activity Promotion Intervention Components on Sedentary Behavior and Light Physical Activity in Breast Cancer Survivors

Abstract

Purpose:

Less sedentary time (ST) and more light physical activity (LPA) are associated with improved health outcomes among breast cancer survivors. However, less is known about whether interventions designed to increase moderate-to-vigorous physical activity (MVPA) also influence time spent sedentary and in LPA. This study explored the overall and component-specific effects of the Fit2Thrive technology-supported MVPA intervention on ST and LPA in breast cancer survivors.

Methods:

Physically inactive breast cancer survivors (n=269; M_age=52.5; SD=9.9) received a core intervention (Fitbit and Fit2Thrive smartphone app) and were randomized to five components (“on” vs. “off”): Support Calls, Deluxe App, Fitbit Buddy, Online Gym, and Text Messages for 12-weeks. Mixed-effects models examined accelerometer-assessed ST and LPA changes, overall, and by intervention component at 12- and 24-weeks. We also examined relationships between changes in MVPA with ST and LPA, and moderating effects of demographic and disease characteristics on component effects.

Results:

Compared to baseline, ST significantly decreased at 12-weeks (β=-14.2 minutes/day; p<0.01); reductions were maintained at 24-weeks (β=-11.0 minutes/day; p<0.01). LPA significantly increased at 24-weeks (β=7.4 minutes/day; p=0.04), but not at 12-weeks. No component effects were significant. Greater increases in MVPA were associated with greater ST decreases and LPA increases. Time since diagnosis moderated effects of the deluxe app on ST and LPA.

Conclusions:

Participation in an mHealth MVPA promotion trial was associated with modest reductions in ST and increases in LPA, while also increasing MVPA. Findings suggest the core intervention may have been sufficient to support favorable changes across movement behaviors. Future research should investigate how to enhance effects while maximizing MVPA and health outcomes.

Implications for Cancer Survivors:

Participation in an mHealth MVPA promotion trial may also result in reduced ST and increased LPA.

INTRODUCTION

There are 4.3 million breast cancer survivors in the United States, with this number expected to increase to 5.3 million by 2035.[1] Moderate-to-vigorous physical activity (MVPA) is consistently associated with reduced risks of chronic diseases (e.g., diabetes and heart disease), cancer mortality, and cancer incidence and recurrence; fewer cancer treatment-related side effects; and improved quality of life among breast cancer survivors.[2–4] It is recommended that cancer survivors engage in MVPA (i.e., ≥150 min/week of moderate- or 75 min of vigorous-intensity activity) and muscle-strengthening physical activity (≥2 times/week).[5–7] However, only 14% of breast cancer survivors meet physical activity guidelines,[8–10] and they spend significantly more time sedentary [i.e., time spent in any waking behavior in a sitting, lying, or reclining position with ≤1.5 metabolic equivalents (METs)][11] and less time active compared to individuals without a history of cancer.[12–14] Data indicate that less than 2% of breast cancer survivors’ waking day is spent in MVPA, while approximately 70% is spent in sedentary time, and approximately 28% in light-intensity physical activity (LPA; behaviors with an energy expenditure of >1.5 to <3 METs).[15] This is particularly concerning because increased sedentary time and reduced light physical activity are independently associated with poorer outcomes among breast cancer survivors.[16–22]

Increased sedentary time is independently associated with a higher risk of all-cause and cancer mortality, with the highest risk associated with sitting ≥8 hours/day.[16,17] Additionally, increased sedentary time is associated with increased fatigue[18,19] and adiposity,[13] and a poorer quality of life[18] in breast cancer survivors. Breast cancer survivors who engage in high levels of sedentary time, even when accounting for MVPA, are at increased risk for higher waist circumference[13,20] and all-cause mortality,[23] as well as increased fatigue,[18] pain,[24] and functional decline.[25] Findings from a meta-analysis reported that the combination of low physical activity and high sedentary time increased the risk of all-cause mortality by 59% (95% CI 1.52, 1.66), and this increased risk was not offset by engaging in the recommended ≥150 minutes/week MVPA.[21] Although increased sedentary time is generally associated with poorer health outcomes, findings from the broader literature also suggest that the magnitude of risk may depend in part on the amount of MVPA accumulated, with some studies showing particularly elevated risk when sedentary time co-occurs with insufficient MVPA.[26–28] Increased LPA is also associated with a lower mortality risk[22] among older adult cancer survivors, independent of MVPA and total activity. Furthermore, for cancer survivors not engaging in MVPA, replacing sedentary time with LPA is associated with improvements in health and mortality risk.[29] Thus, increased PA (MVPA and LPA) and reduced sedentary time may maximize health outcomes among breast cancer survivors.

Many studies have intervened to increase MVPA among cancer survivors,[30–32] but few have assessed how these interventions influence sedentary time[33–40] and/or LPA.[41] Understanding how MVPA interventions influence sedentary time and LPA is important because time is finite. An increase in MVPA may displace time spent in sedentary behavior or LPA, each of which has distinct health implications. Relationships among these behaviors may be particularly relevant for cancer survivors who experience fatigue, mobility limitations, or other treatment-related side effects,[18,19,24,25] which may impact how time is spent when MVPA is increased.

Increasing MVPA could have null (i.e., no change) or negative (i.e., increased sedentary time and/or reduced LPA) effects on sedentary time and LPA, as survivors may compensate for increased energy expenditure or feelings of exertion by reducing other activities or sitting/resting more. For example, Pinto et al.[42] reported unchanged sedentary time among breast cancer survivors who participated in an exercise intervention and increased their MVPA time. However, increased MVPA could also lead to positive changes in these behaviors (i.e., decreased sedentary time and/or increased LPA) because individuals may be motivated to make additional behavioral changes, and some strategies used to increase MVPA could positively affect these other behaviors.[35,36,43–47]

A nuanced understanding of the relationships among MVPA, sedentary time, and LPA can inform the development of interventions that promote MVPA without unintended negative effects on sedentary time or LPA. To date, findings from studies assessing changes in sedentary time and LPA as part of interventions promoting MVPA among cancer survivors have been heterogeneous, with some reporting positive effects and others negative effects.[35–37,48] There are several potential explanations for these varying effects, including small sample sizes (ranging from 30 to 87 participants), varying amounts of in-person contact, and short intervention periods (ranging from 8 to 13 weeks), with the exception of one study (24 weeks).[37] In addition, these studies all tested multicomponent interventions compared to a control using a randomized controlled trial (RCT) design,[35–37,48] which makes it difficult to understand whether specific intervention components may differentially influence sedentary time and LPA.

The primary purpose of this study is to conduct a secondary analysis of Fit2Thrive, a full-factorial mHealth intervention that resulted in significant increases in MVPA,[49] to explore the overall and intervention component-specific effects on accelerometer-assessed sedentary time and LPA at 12 and 24 weeks.

METHODS

Study Design

The Fit2Thrive study was a full-factorial (5²) randomized controlled trial that used the Multiphase Optimization Strategy (MOST) methodology.[50] Full details of the study protocol are described elsewhere.[51] The primary purpose of Fit2Thrive was to examine the effects of five technology-supported intervention components, each with two levels (“on” vs. “off”), on weekly MVPA minutes at 12 and 24 weeks. Overall, participants increased MVPA significantly post-intervention (12 weeks), with effects maintained but attenuated at the 24-week follow-up.[49] However, no significant effects were observed for any components “on” vs. “off”.

Participants

Participants were recruited using an email listserv that connected researchers with individuals interested in participating in breast cancer research (Love Research Army©). Inclusion criteria included being female, ≥18 years, diagnosis with stage I to III breast cancer within 5 years, completion of primary treatment (surgery, chemotherapy, or radiation therapy) >3 months prior, no major contraindications to MVPA, ability to read, write, and speak English, smartphone ownership, having computer and internet access, willingness to receive phone calls and find a buddy to participate with them (if condition is assigned), and self-reporting <60 minutes of weekly MVPA. Individuals who screened at high risk for a cardiovascular event, as indicated by the Physical Activity Readiness Questionnaire (PAR-Q),[52] were required to obtain medical clearance. All participants completed informed consent before engaging in any study activities.

Randomization

After completing baseline assessments, participants were randomized using computer-generated, randomly permuted blocks to one of the 32 potential conditions, representing all possible combinations of the five intervention components.

Fit2Thrive Intervention Components

All participants received the core intervention and were randomized to receive some combination of none to five of the intervention components: Support Calls, Deluxe App, Fitbit Buddy, Online Gym, and Text Messages. Intervention components tested were selected based on prior formative work with breast cancer survivors.[53,54] Social cognitive theory guided the core and intervention components.[55]

A detailed description of the intervention is provided elsewhere.[49,51,56] However, a brief overview of components is highlighted here. The core intervention consisted of a Fitbit Zip and the custom-built Fit2Thrive MVPA self-monitoring smartphone app.[57] Participants received an exercise prescription to gradually increase MVPA to 150 minutes/week. Educational information was also provided on the benefits of physical activity and on approaches to encourage engagement and maintenance of physical activity (e.g., goal setting, social support). Participants assigned to telephone Support Calls (i.e., “on” level) received bi-weekly semi-scripted coaching calls (6 total calls) from a study team member. Calls reviewed progress toward the MVPA goal and provided support for goal-setting, problem-solving, and building self-efficacy. When the Deluxe App was “on,” participants received additional app features including a weekly goal-setting tool, a weekly challenges tool, and a study newsfeed. Additional details of the tools used are published elsewhere. Those assigned to the Fitbit Buddy component “on” were asked to choose an individual from their personal life to be their “buddy” during the trial. Selected buddies received a Fitbit device and educational content about providing social support. Automated, biweekly emails with information and strategies to enhance social support were sent to participants and their buddies via REDCap. Individuals assigned to the Online Gym component as “on” received unlimited access to an online exercise video platform. Additionally, automated emails with suggested weekly exercise schedules were sent to participants via REDCap. Those assigned to the Text Messaging component “on” received 1 to 5 daily tailored app notifications that included motivational messages, progress feedback on MVPA, and educational content offering simple ways to increase activity. Variation in the number of messages per day resulted from when event-contingent messages (e.g., progress notifications, reminders to sync) were sent, as well as those on a fixed schedule (e.g., motivational messages, educational content). Additional details on message logic can be found elsewhere.[49,51]

Adherence to the components varied, and details on feasibility and acceptability are described elsewhere.[56] Briefly, core intervention adherence was high: 98% of participants downloaded the Fit2Thrive app and connected their Fitbit; the Fit2Thrive app was opened an average of 4.7 times/week during the 12-week intervention, and participants had valid Fitbit data on 93% of study days.[56] Adherence rates by component were as follows: 99% of Support Calls were attended; Deluxe App features ranged from 28% (daily goal setting tool) to 39% (weekly challenge tool); 89% of buddies connected their Fitbit and 88% of participants reported discussing Fit2Thrive with their buddy ≥1 time per week; 55% of participants reported using the Online Gym ≥1 day/week and 41% of Text Messages were marked as read.

Measures

Participants completed demographic and health history questionnaires online via REDCap[58,59] at baseline. Sedentary time, LPA, and MVPA were measured objectively via accelerometry at baseline, 12-weeks, and 24-weeks.

Demographics and Health History

Participants self-reported age, race, ethnicity, education, income, medical and cancer history (i.e., cancer diagnosis and treatment, chronic conditions), and height and weight to calculate BMI (kg/m²).

Sedentary Time and LPA

Daily minutes of sedentary time and LPA were assessed using the ActiGraph GT3X-BT accelerometer (ActiGraph, Pensacola, Florida).[60,61] During assessment periods, participants were instructed to wear the accelerometer on an elastic waistband on their non-dominant hip for seven consecutive days during waking hours and to complete a wear log. Following their 7-day assessment, they returned the accelerometer and the log using the provided postage-paid return envelope.

Accelerometer data were downloaded, and 60-second epoch data were processed with ActiLife 6.0 software.[62–65] using established cut points to classify every wear minute as sedentary, light, moderate, or vigorous. Sedentary time was defined as <100 counts/min, LPA as 100–1951 counts/min, and MVPA as ≥1952 counts/min.[64] Intervals of ≥90 minutes of zero vertical axis activity counts were considered non-wear.[66] This allowed for 2-minute interruptions (>0 counts per minute) with upstream and downstream screening for artifactual movements, totaling 30 minutes.[67] Self-report wear logs were used to verify the number of wear days included.[60] A day was considered valid and included if the wear time was ≥10 hours.[68,69] Participants were required to have at least 3 valid days of wear to be included in analyses at each time point.

Statistical Analysis

Intent-to-treat analyses, using separate linear mixed-effects models with a random intercept, examined the effects of intervention components on changes in mean daily minutes of sedentary time and LPA at 12 and 24 weeks. For the five experimental components, effect coding was used, with “off” levels coded as −1 and “on” levels coded as 1. Indicator variables for 12 and 24 weeks, as well as interactions between time indicators and the treatment components, were included in the models. Overall mean changes from baseline, along with treatment effects (mean differences in baseline to follow-up changes between “on” and “off” levels) and their corresponding 95% confidence intervals (CIs), were calculated at 12 and 24 weeks. To assess whether changes in mean daily MVPA from baseline were associated with mean daily sedentary time and LPA at follow-up, we fitted separate linear mixed models adjusting for baseline MVPA and wear time. Finally, we conducted a post hoc, exploratory moderator analysis to examine whether baseline characteristics, including age (<60 vs. ≥60), BMI (<25 vs. ≥25), time since diagnosis (<30 months vs. ≥30 months), education (<bachelor's degree vs. ≥ bachelor's degree), comorbidities (0 vs. ≥1), chemotherapy (yes vs. no), and radiation therapy (yes vs. no) influenced the effects of intervention components on changes in daily sedentary and LPA minutes. All models were adjusted for wear time continuously.

Statistical analyses were conducted with SAS 9.4 (SAS Institute, Inc., Cary, North Carolina).[70] Statistical significance was defined by a p-value <0.05. For moderator analyses, a Bonferroni correction was applied to maintain the family-wise error rate at 0.05, corresponding to a significance threshold of p<0.007.

RESULTS

Participants

A total of 269 participants were randomized. Valid accelerometer data were obtained from a total of 266 participants at baseline, 246 at 12 weeks, and 228 at 24 weeks. One participant was excluded from regression analyses because of implausible MVPA values at 12 and 24 weeks, and two participants lacked baseline data due to staff error. On average, participants were 52.5±9.9 years, 87% White, 93.3% non-Hispanic, with a 28.9 kg/m² body mass index, and 3.3±2.4 years since breast cancer diagnosis. See Table 1 for participant baseline characteristics. Baseline characteristics, daily sedentary time, and LPA minutes did not differ significantly across intervention components.

Table 1.

Baseline participant characteristics.

	Total (n=268)
Age (years)	52.5 (9.9)
Body mass index (kg/m2)	28.8 (6.2)
Time Since Diagnosis (years)	3.0 (2.4)
>30 months (count)	136 (50.8%)
Ethnicity
Hispanic/Latino	18 (6.7%)
Race
White	233 (86.9%)
Black or African American	16 (6.0%)
Other	19 (7.1%)
Daily sedentary time (minutes)	602.9 (89.5)
Daily light physical activity (minutes)	249.0 (64.5)
Daily moderate-vigorous physical activity (minutes)	14.1 (12.2)

Data for age, body mass index, time since diagnosis, sedentary and light physical activity time are mean (SD); data for time since diagnosis >30 months, ethnicity, and race are n (%).

Sedentary Time

When estimating the mean overall effects, sedentary time significantly decreased by an average of 14.2 minutes/day (95% CI [−21.1, −7.2], p<0.001) from baseline to 12-weeks. These reductions were maintained at 24-weeks [−11.0 minutes/day (95% CI [18.1, −3.8], p=0.0031)]. All 2-way interactions between each component and time (conceptually equivalent to component main effects) were not significant for daily sedentary time (see Table 2). Changes in daily MVPA were significantly associated with changes in sedentary time, while adjusting for wear time and baseline MVPA. For every 1-minute increase in MVPA from baseline, there were 1.5 fewer minutes of sedentary time at follow-up (95% CI [−1.8, −1.2], p<0.0001).

Table 2.

Predicted mean change in sedentary behavior (minutes/day) from baseline by experimental components.

Experimental Component	Week	Mean Change M (SE)		On vs. Off Mean Change Difference (95% CI)	p-value
		ON	OFF
Support Calls	12	−18.0 (4.8) *	−10.3 (4.8) *	−7.7 (−20.5, 5.0)	0.23
	24	−12.3 (4.9) *	−9.7 (5.0)	−2.7 (−15.9, 10.5)	0.69
Deluxe App	12	−14.1 (4.8) *	−14.2 (4.9) *	0.2 (−12.6, 12.9)	0.98
	24	−11.9 (4.9) *	−10.0 (5.0) *	−1.9 (−15.1, 11.3)	0.78
Buddy	12	−15.8 (4.8) *	−12.5 (4.9) *	−3.4 (−16.1, 9.4)	0.61
	24	−13.7 (4.9) *	−8.3 (5.0)	−5.4 (−18.6, 7.8)	0.42
Online Gym	12	−10.3 (4.8) *	−18.1 (4.8) *	7.8 (−5.0, 20.6)	0.23
	24	−11.2 (4.9) *	−10.8 (5.0) *	−0.4 (−13.6, 12.8)	0.95
Text Messages	12	−20.3 (4.8) *	−8.0 (4.8)	−12.3 (−25.1, 0.5)	0.06
	24	−9.4 (5.0)	−12.6 (4.9) *	3.2 (−10.0, 16.4)	0.64

Note.

^*=value indicates statistical significance from baseline at alpha=0.05 level; data are based on mixed linear regression controlling for accelerometer wear time. Abbreviations: SE=standard error; CI=confidence intervals.

In exploratory analyses, among participants who were ≤30 months post-diagnosis, those with the Deluxe component “off” reduced their sedentary time by 20.5 minutes more than those with the Deluxe component “on” at 24-weeks. Additionally, among participants >30 months post-diagnosis, those with the Deluxe component “off” increased their sedentary time by 24.2 minutes compared with those with the Deluxe component “on”. No other moderation effects were significant.

LPA Results

When estimating the mean overall effects, daily LPA increased from baseline to 12-weeks by 6.3 minutes/day (95% CI [-0.1, 12.7], p=0.05), but these changes were not significant. However, daily LPA significantly increased by 7.4 minutes/day (95% CI [0.5, 14.4], p=0.03) at 24-weeks. There were no significant main effects on LPA for any of the 5 components “on” vs. “off” at 12- or 24-weeks (see Table 3). Changes from baseline to 24-weeks for MVPA were significantly associated with LPA, adjusting for wear time and baseline MVPA. Every 1-minute increase in MVPA from baseline was associated with an additional 0.5 minute (95% CI [0.2, 0.8], p<0.001) in LPA at follow-up.

Table 3.

Predicted mean change in light-physical activity (minutes/day) from baseline by experimental components.

Experimental Component	Week	Mean Change M (SE)		On vs. Off Mean Change Difference (95% CI)	p-value
		ON	OFF
Support Calls	12	9.7 (4.4) *	3.0 (4.5)	6.7 (−5.4, 18.7)	0.28
	24	8.5 (4.8)	6.5 (4.9)	2.0 (−11.0, 15.1)	0.76
Deluxe App	12	7.4 (4.4)	5.3 (4.5)	2.1 (−10.0, 14.2)	0.73
	24	9.7 (4.8) *	5.3 (4.9)	4.5 (−8.6, 17.5)	0.50
Buddy	12	8.2 (4.4)	4.5 (4.5)	3.7 (−8.4, 15.7)	0.55
	24	10.2 (4.8) *	4.8 (4.9)	5.4 (−7.7, 18.4)	0.42
Online Gym	12	3.5 (4.5)	9.2 (4.5) *	−5.7 (−17.7, 6.4)	0.35
	24	9.2 (4.8)	5.8 (4.9)	3.5 (−9.6, 16.5)	0.60
Text Messages	12	11.6 (4.5) *	1.1 (4.4)	10.5 (−1.5, 22.6)	0.09
	24	5.7 (4.9)	9.3 (4.8)	−3.7 (−16.7, 9.4)	0.58

Note.

^*=value indicates statistical significance from baseline at alpha=0.05 level; data are based on mixed linear regression controlling for accelerometer wear time. Abbreviations: SE=standard error; CI=confidence intervals.

Considering participants who were ≤30 months post-diagnosis, those with the Deluxe component “off” increased their LPA by 16.0 minutes more than those with the Deluxe component “on” at 24-weeks. In addition, among participants >30 months post-diagnosis, those with the Deluxe component “off” increased their LPA by 25.4 minutes compared with those with the Deluxe component “on”. No other moderator effects were significant for LPA at 12-weeks or 24-weeks.

DISCUSSION

Overall, our findings indicate that participation in the Fit2Thrive MVPA intervention was associated with modest reductions in sedentary time at 12 weeks, which were maintained at 24 weeks, and modest increases in LPA at 24-weeks. No significant differences were observed between “on” and “off” levels for sedentary time or LPA across any intervention component. Increased MVPA from baseline to follow-up was associated with less sedentary time and more LPA at follow-up. Collectively, these findings suggest that an mHealth intervention designed to increase MVPA may also produce small favorable changes in other movement behaviors, even when sedentary time and LPA are not directly targeted.

Our findings regarding the effects of Fit2Thrive on sedentary time are consistent with Weiner et al.[36] who found that, among breast cancer survivors, a 12-week remotely delivered, Fitbit-based MVPA intervention significantly reduced sedentary time (-24.9 minutes/day) compared with controls, and that each 1-minute increase in MVPA was associated with a ~2-minute decrease in sedentary time. In contrast to Fit2Thrive, where we did not see a significant increase in LPA until week 24, LPA significantly increased (+11.4 minutes/day) in the intervention group at 12-weeks in Weiner et al.,[36] but between-group differences were not significant. Significant moderating effects of time since surgery on sedentary time were also observed in that study, similar to the effects we observed for time since diagnosis on deluxe app effects. The reductions in sedentary time and LPA in Fit2Thrive were smaller in magnitude than those of Weiner et al.[36] which may be attributed to the fact that participants in that study increased MVPA by ~45 minutes per week (99.4 minutes vs. 53.6 in Fit2Thrive [49]) and had lower sedentary time and LPA at intervention start, suggesting they may have been more able and motivated to change these behaviors. Our findings are in contrast to two other studies in breast cancer survivors, which observed no change in sedentary time as a result of physical activity intervention participation,[42,71] and Blair et al,[35] which found no significant changes in sedentary time or LPA when specifically targeting the replacement of sedentary time with LPA among older cancer survivors.

Furthermore, two recent meta-analyses found reductions in sedentary time only for interventions focused solely on reducing sedentary behavior, not for those targeting increased MVPA or both increased MVPA and reduced sedentary time in non-cancer populations,[72,73] suggesting the importance of behavior specificity. As such, it is not surprising that the change in sedentary time observed in Fit2Thrive was also smaller than that observed in Lynch et al.[33] (−37 minutes/day sedentary time), which targeted both reducing sedentary time and increasing MVPA among breast cancer survivors, and a meta-analysis of sedentary behavior reduction interventions in chronic disease populations (sedentary time mean difference= −30.80 minutes/day). While the observed reduction in sedentary time in Fit2Thrive (14 minutes/day) falls below commonly cited thresholds, such as 30 minutes/day,[74] and participants remained highly sedentary overall, the direction of change is notable, given that sedentary time was not explicitly targeted. Therefore, these findings should not be interpreted as indicating that Fit2Thrive meaningfully resolved high sedentary time, but rather that the intervention may have yielded a modest ancillary benefit on sedentary behavior.

There are several reasons we may have observed a positive effect on sedentary time and LPA in Fit2Thrive without specifically targeting these behaviors. First, our sample was recruited via a listserv for women interested in breast cancer research who may have been more motivated to improve their health, and applied the MVPA behavior change strategies from Fit2Thrive to reduce sedentary time and increase light-intensity activity. Second, our study, as well as Lynch et al.[33] and Weiner et al.,[36] used wearable devices with features that increase awareness of sedentary time and LPA. Recent meta-analyses examining device-based interventions among adults have demonstrated significant increases in steps and MVPA and decreases in sedentary behavior, with wearables used to promote MVPA.[75,76] Notably, the reduction in sedentary time (14 min/day) was modest, and participants remained highly sedentary overall. Nevertheless, the direction of change is encouraging given that sedentary time was not explicitly targeted, and even modest reductions may be relevant when delivered through scalable, low-burden interventions. Future research is warranted to better understand which intervention characteristics yield the most favorable changes in MVPA, sedentary time, and LPA. Additionally, understanding how to enhance effects on sedentary time and LPA within MVPA mHealth interventions to optimize health outcomes without overburdening participants or compromising intervention effects on MVPA is needed.

No significant differences were observed between the ‘on’ and ‘off’ levels of any individual intervention component for sedentary time or LPA. Several factors may explain this finding. Given that all participants received the core intervention, the core package itself may have been sufficient to produce favorable changes in these behaviors, leaving little room for the additive effects of individual components. In addition, the components were designed primarily to optimize MVPA rather than to target sedentary time or LPA specifically, and small component-specific effects may therefore have been difficult to detect. This interpretation is consistent with the primary outcome paper, which also did not identify significant main effects of individual components on MVPA. Additionally, while the intervention was designed to promote MVPA, several intervention features may also have indirectly addressed determinants of sedentary behavior highly relevant to cancer survivors, including fatigue, self-efficacy, and motivational barriers. Self-monitoring, goal setting, feedback, and supportive coaching may have helped participants build confidence to move more throughout the day, not only during planned bouts of MVPA, which may partly explain the observed reductions in sedentary time and later increases in LPA.

Time since diagnosis emerged as a moderator of the Deluxe App. Although this finding should be interpreted cautiously, particularly in the absence of a significant main effect, it raises the possibility that survivors earlier versus later in recovery may differ in symptom burden, routine stability, and responsiveness to intervention features. Potentially reflecting heterogeneity in response to added intervention complexity by time since diagnosis or a chance finding requiring replication. Future studies should prospectively examine time since diagnosis as a potential moderator when determining the complexity and timing of intervention support needed to influence sedentary time and LPA.

Strengths and limitations

This work has several strengths. First, this study was conducted fully remotely, which reduced geographic and in-person participation barriers and enabled a broader reach than an in-person trial. Second, the use of a factorial experiment allows for testing of the intervention components and their influence on changes in sedentary time and LPA. Third, our sample size (n=269) was relatively large compared to similar studies of breast cancer survivors (n=85). Finally, we used objective measures of both sedentary behavior and LPA.

However, this study is not without limitations. The generalizability of these findings may be limited due to the lack of heterogeneity, with most of the sample being White, non-Hispanic, middle-class, and early-stage breast cancer survivors. This study was also conducted fully remotely, which reduced geographic and in-person participation barriers and, in principle, allowed for enrollment of eligible participants from across the United States. However, the study's reach was constrained by recruitment through the Love Research Army© listserv, which likely attracted individuals already interested in breast cancer research who may not be representative of the broader population of breast cancer survivors. Although individuals were ineligible if they self-reported >60 minutes/week of MVPA at baseline, more than half of the sample exceeded this threshold with device-measured MVPA.[49] This discrepancy may reflect challenges in accurately recalling or classifying moderate-intensity activity in daily life, although underreporting of activity to meet eligibility criteria also cannot be ruled out. Thus, findings may not apply to less active and unmotivated populations. In addition, because sedentary time and LPA were secondary outcomes and the trial did not specifically target these behaviors, small but potentially meaningful component-specific effects may have gone undetected.

The core wearable in the Fit2Thrive intervention was the Fitbit Zip, which was practical and acceptable at the time of trial design but offered fewer real-time prompting and self-monitoring features than newer wrist-worn devices. Devices that provide more immediate feedback and movement prompts may be especially relevant for reducing sedentary time and increasing LPA, and future work should examine how engagement with these features influences outcomes. Additionally, using a hip-worn accelerometer to measure sedentary behavior may overestimate sedentary time, as estimates are based on lack of movement and cannot distinguish between standing and seated positions.[77] Future studies should consider using thigh-worn devices, such as the activPAL, to measure sedentary behavior. We did not examine baseline MVPA as a moderator of the component effects. Future analyses should examine whether intervention effects on sedentary time and LPA differ by baseline activity status. Finally, the follow-up period was only 24-weeks, which may limit our understanding of longer-term effects on sedentary time and LPA, and we cannot determine whether the modest changes observed would translate into clinically meaningful long-term benefit.

Conclusions

Our findings indicate that sedentary time decreased and LPA increased modestly among breast cancer survivors participating in a mHealth MVPA promotion trial. These secondary findings suggest that the Fit2Thrive core intervention may have supported small, favorable changes in movement behaviors beyond MVPA and no individual intervention component were superior. Future work is warranted to identify how to incorporate behavior-specific and resource-efficient strategies to more effectively reduce sedentary time and increase LPA without compromising MVPA outcomes or overburdening cancer survivors.

Data Availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.