Long-Term Engagement With a Physical Activity App Among Gynecologic Cancer Survivors in the LETSGO Trial: 12-Month Prospective, Multicenter, Quasi-Experimental Study

Background: Physical activity (PA) alleviates many treatment-related side effects in gynecologic cancer survivors, yet long-term PA levels remain low. Mobile health interventions can support self-management and increase PA levels; however, evidence from real-world, year-long engagement with smartphone apps in this population is still limited. Objective: The aim of this study is to describe 12-month user engagement with the PA component of a smartphone app implemented within a partially nurse-led routine follow-up in a real-world cohort of gynecologic cancer survivors. Methods: This descriptive study analyzed server-generated log data from the intervention arm of the prospective, multicenter, quasi-experimental LETSGO (Lifestyle and Empowerment Techniques in Survivorship of Gynecologic Oncology) trial (NCT04122235). Between December 2019 and July 2022, 378 cancer survivors (ovarian, endometrial, cervical, vulvar, or vaginal cancer) from 5 Norwegian hospitals were enrolled in the intervention arm and were offered the app plus a Garmin Vivofit 4 activity tracker alongside standard consultations. Primary outcomes for this study were (1) weekly PA registrations (objective step counts from the activity tracker and self-reported PAs) and (2) temporal patterns of step logging during each cancer survivor's first 52 weeks postenrollment. Secondary analyses compared baseline characteristics of app users (≥2 wk of PA logging) and nonusers. Results: Of 378 eligible participants (mean age 63, SD 13 y; BMI 28.5, SD 6.4 kg/m²), 272 (72%) logged at least 2 weeks of PA, and 225 (60%) synchronized objective step data. Mean daily steps were 5657 (SD 2799; median 5533, IQR 3499-7498). Step tracking dominated app use (mean 21, SD 16 logged wk), followed by self-reported walking (mean 23, SD 19 wk) and resistance training (mean 15, SD 16 wk). Weekly step-logger counts fell around study weeks 16, 32, and 44, but rebounded by 30 to 50 participants within 3 weeks, indicating episodic rather than permanent disengagement. App users were younger (mean difference -6.7 y, 95% CI -9.6 to -3.9; P<.001), more often employed (χ²1=12.7; P<.001), and more likely to have higher education (χ²2=13.6; P<.001) than nonusers. Tumor type and treatment modality were not associated with engagement. Conclusions: In a routine-care setting, nearly three-quarters of gynecologic cancer survivors engaged repeatedly with an app-supported PA module over 12 months, although mean PA levels were modest and participation was more prevalent among younger, employed, and more highly educated participants. Engagement followed an ebb-and-flow pattern, suggesting that built-in re-engagement prompts and equity-focused onboarding are needed to sustain and broaden participation. These findings support the feasibility of blended mobile health follow-up while highlighting the importance of adaptive strategies to promote long-term PA adherence and bridge the digital divide among cancer survivors.