Feasibility of an exercise intervention for chemotherapy-related fatigue in patients with gynecologic cancer: a randomized controlled trial with multidimensional assessment Authors: Natsuko Kamiya, Yuichi Imai, Yusuke Saigusa, Risa Matsunaga, Aiko Kawano, Tae Mogami, Kanna Takagi, Kazuko Hayashi, Kae Tachibana, Fujiko Sakuma, Tatsuya Matsunaga, Manabu Nonogaki, Takeshi Nakamura, Etsuko Miyagi, Taichi Mizushima Published: 2026 ABSTRACT BackgroundAlthough exercise interventions have shown benefits in several cancer populations, evidence during chemotherapy for gynecologic cancers remains limited. This randomized controlled trial evaluated the feasibility and efficacy of a home-based exercise intervention for cancer-related fatigue, quality of life, and physical function in patients undergoing chemotherapy for gynecologic cancers.MethodsA total of 110 patients with cervical, endometrial, or ovarian cancer were randomized 1:1 to an exercise or control group during paclitaxel–carboplatin chemotherapy. The home-based intervention comprised 30 min of moderate-intensity walking and squat exercises three times weekly. Participants attended rehabilitation visits every 3 weeks for feedback on safety, appropriate technique, and adherence. Control participants received standard care without exercise instructions but recorded their daily activity. The primary endpoint was change in fatigue measured using the Functional Assessment of Chronic Illness Therapy–Fatigue (FACIT-F) subscale. Secondary endpoints included other patient-reported outcomes, physical function, body composition, and adverse events.ResultsFifty-one and 55 patients in the exercise and control groups, respectively, were evaluable. No significant difference in FACIT-F change was observed (LS mean difference: 1.8; 95% CI: − 1.3 to 4.9). Exercise adherence was 76% in the exercise group; 71% of the controls voluntarily exercised at similar levels. Patients with prior exercise habits were more likely to meet exercise criteria (81% vs. 65%; p = 0.07). The exercise group showed better preservation of quadriceps strength; CT revealed reduced psoas muscle mass and increased visceral fat in both groups. Grade ≥ 2 nausea and myalgia were less frequent in the exercise group.ConclusionsAlthough the intervention did not significantly improve fatigue, it was safe, feasible, and associated with preserved lower limb strength and fewer non-hematologic toxicities. High adherence in both groups suggests that exercise during chemotherapy is achievable and well accepted, reflecting strong motivation to remain active, which may also have contributed to control-group contamination and reduced between-group differences. Future studies should refine exercise protocols through digital monitoring and combined nutritional strategies while identifying patient subgroups most likely to benefit during active cancer treatment.Trial registrationRegistered with the Japan Registry of Clinical Trials (jRCT; registration number: jRCT1031200069; date: July 16, 2020).Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s13102-026-02001-3. Background In recent years, patient-reported outcomes (PROs), particularly those related to quality of life (QoL), have become an integral component of clinical oncology trials. Monitoring QoL during chemotherapy, along with the implementation of individualized supportive care, is increasingly recognized as essential to comprehensive cancer management. Among patients undergoing chemotherapy for gynecologic cancers, fatigue is one of the most common and debilitating symptoms, affecting approximately 70–90% of patients [1]. Despite its high prevalence and clinical relevance, no standardized pharmacologic intervention is currently available for chemotherapy-induced fatigue, unlike established treatments for nausea or pain. Emerging evidence from clinical studies involving patients with breast, colorectal, and hematologic malignancies suggests that exercise interventions delivered concurrently with chemotherapy can reduce cancer-related fatigue and improve overall QoL [2, 3]. Similar outcomes have been observed in prostate cancer populations undergoing androgen deprivation therapy, where exercise appears to mitigate treatment-related fatigue [4]. Randomized controlled trials (RCTs) in breast cancer cohorts further support the efficacy of structured exercise programs in improving not only fatigue but also psychological health, myelosuppression, and treatment-related gastrointestinal symptoms, such as nausea [5–8]. These findings underscore the potential role of exercise as a non-pharmacological modality in mitigating chemotherapy-related toxicities and enhancing functional outcomes during active cancer treatment. However, research data on gynecologic cancers remain relatively limited. RCTs of patients with gynecologic cancers are scarce, often having small sample sizes or focusing on cancer survivors post-treatment rather than patients undergoing chemotherapy [9–13]. Among these RCTs, some have suggested potential benefits, such as improvements in psychological well-being in endometrial cancer survivors through aerobic exercise and fatigue reduction in patients with ovarian cancer through exercise interventions guided by nurses and incorporating cognitive behavioral therapy (CBT). Nevertheless, evidence regarding the efficacy of exercise interventions in managing chemotherapy-induced fatigue in patients with gynecologic cancer remains limited. Another important consideration is age. Compared to breast cancer trials, which typically include participants aged 45–60 years, patients with gynecologic cancers—particularly ovarian or endometrial cancer—tend to be older, often in their 50s, 60s, or beyond. This age-related difference was one of the motivations for the present study, as the effectiveness and feasibility of exercise interventions may differ across age groups. In the present study, we conducted a prospective RCT involving approximately 100 patients and used PROs to investigate whether exercise therapy, compared with usual care, improves fatigue during chemotherapy. In addition, we assessed changes in physical function, including lower limb muscle strength, and body composition before and after chemotherapy using computed tomography (CT). Methods Study design and participants This phase 2 prospective RCT was conducted at Yokohama City University Hospital and Yokohama City University Medical Center. The study was approved by the Ethics Committee of Yokohama City University (approval number: F241200001) and registered in the Japan Registry of Clinical Trials (jRCT1031200069, registration date: July 16, 2020). A total of 110 patients with gynecologic cancers (cervical, endometrial, and ovarian cancer) were randomized 1:1 to either an exercise intervention group or a control group, stratified by study site, performance status, and cancer type. Participants were enrolled by the principal investigators at each site, and randomization was performed using minimization within the UMIN Internet Data and Information system for Clinical and Epidemiological research, Cloud version (INDICE cloud) [14]. This web-based central randomization system ensured allocation concealment. Enrollment occurred between July 2020 and August 2024 according to the inclusion and exclusion criteria. The inclusion criteria were as follows: (1) age 20–79 years; (2) first-time conventional paclitaxel–carboplatin (TC) therapy, either initial or recurrent (excluding neoadjuvant); (3) Eastern Cooperative Oncology Group (ECOG) performance status 0–2; (4) ability to complete questionnaires and assessments; and (5) provision of written informed consent. The exclusion criteria included the following: (1) requirement for rehabilitation before treatment due to comorbidities or poor general condition; (2) completion of ≥ 2 cycles of chemotherapy unlikely due to cancer-related complications or progression; (3) concurrent malignancies requiring treatment (remission cases were eligible); (4) ongoing depression or sleep disorders requiring regular medication; and (5) severe cardiovascular disease, including NYHA class ≥ III with palpitations, dyspnea, or chest pain. Study intervention The chemotherapy schedule and hospital visit frequency were the same in both groups. All patients were required to visit the gynecology department on the day of chemotherapy and again 2 weeks later for routine follow-up. Patients were asked to record their exercise volume in a diary. Control patients received no exercise guidance but were asked to complete the diary and record their exercise volume at home. In Japan, cancer rehabilitation is primarily provided to inpatients, and exercise instruction is not routinely included in standard care for patients receiving outpatient chemotherapy. For the exercise intervention group, during the chemotherapy treatment period, patients were instructed to perform a home-based exercise program. The exercise program was designed according to the FITT principle (frequency, intensity, time, and type), informed by established guidelines and previous studies. The World Health Organization recommends that adults with chronic conditions, including cancer survivors, engage in at least 150–300 min of moderate-intensity aerobic physical activity (or 75–150 min of vigorous activity) per week, along with muscle-strengthening activities for ≥ 2 days per week [15]. The American College of Sports Medicine further reports that moderate-intensity aerobic exercise performed ≥ 3 times per week, including sessions of ≥ 30 min and continued for ≥ 12 weeks, is effective in reducing cancer-related fatigue, and that combining aerobic and resistance training is beneficial [16]. As these recommendations are primarily derived from date involving cancer survivors, we additionally considered prior interventional studies in patients undergoing active treatment when determining exercise prescription [7, 9, 17, 18]. Accordingly, we implemented a home-based program consisting of moderate-intensity walking (Borg scale 11–13) for 30 min, three times per week, combined with squat-based resistance training. Given the higher risk of peripheral neuropathy and joint pain associated with the chemotherapy regimen used in this study, resistance training was prescribed as a low-intensity exercise. The squat exercise was performed as a light, comfortable movement, allowing the knees to move forward if tolerated, with supportive actions, such as touching the wall, to maintain balance and safety. The number of repetitions and sets was individually adjusted according to each patient’s physical condition and any complications, as determined by physical therapists and rehabilitation physicians. In the exercise intervention group, exercise adherence and safety were reviewed at the gynecology outpatient visit on each chemotherapy day using the activity diary. Two weeks after each cycle, patients attended the gynecology visit and additionally visited the rehabilitation department, where they received exercise guidance and 10–15 min of supervised training. The activity diary included records of the duration of walking performed at a Borg scale intensity of 11–13 (or whether no exercise was performed), the number of squat repetitions, and any additional physical activity, including the type and intensity. Additionally, patients were able to record their physical condition and symptoms. For this study, adherence was evaluated based on whether each participant’s average weekly walking duration over the entire intervention period reached the prescribed target level (30 min/session, three sessions/week at a Borg scale of 11–13; ≥90 min/week). Squat exercises were not included in the adherence assessment because the number of repetitions and sets was individualized according to each participant’s physical condition and therefore did not have a uniform target across participants. The adherence rate was calculated as the proportion of participants who met this target. The intervention was designed to maintain the prescribed exercise level throughout chemotherapy rather than progressively increasing exercise intensity or volume. If patients experienced poor health during treatment and were unable to continue the prescribed exercises, they were allowed to reduce their exercise volume. Although the prescribed exercise was 30 min per session, three times per week, participants who were unable to complete a full 30-minute session were allowed to divide each session into 20 and 10 min, provided that the total weekly exercise volume remained unchanged. As this does not change the total exercise volume, it was not considered a reduction or discontinuation of prescribed exercise. The intervention was designed to maintain the prescribed exercise level throughout chemotherapy rather than progressively increasing exercise intensity or volume. Chemotherapy Paclitaxel was administered at 175 mg/m² and carboplatin at AUC 6 every 3 weeks. Dose reductions or treatment delays were permitted at the discretion of the attending physician. The use of bevacizumab in combination with ovarian and cervical cancer treatment was allowed. Outcomes The primary endpoint was defined as the change in fatigue, assessed using the Functional Assessment of Chronic Illness Therapy–Fatigue (FACIT-F) subscale (score range: 0–52) [19], from baseline to the end of chemotherapy. Secondary endpoints included FACIT-F scores during treatment, other PROs, physical function assessments, and body composition measured by CT. FACIT-F scores were assessed at the end of cycles 2, 4, and 6. Additional PROs included the Functional Assessment of Cancer Therapy–General (FACT-G) [20], the FACT/GOG-Neurotoxicity (Ntx) subscale [21], and the Pittsburgh Sleep Quality Index (PSQI) [22]. PROs were completed on paper forms and sealed by the patients themselves before being submitted in envelopes to the administrative staff at the reception desk, ensuring that responses were not accessible to the treating physicians. Physical function was evaluated before and after chemotherapy using manual muscle testing (MMT), handgrip strength, quadriceps strength measured with a handheld dynamometer, the Five Times Sit-to-Stand (5STS) test, and the 6-min walk test (6MWT). Handgrip strength was evaluated using a dynamometer, and the highest value from the repeated measurements was recorded. Quadriceps strength was measured using a handheld dynamometer in a standardized seated position. The 5STS test was performed on a standard chair without armrests, and the time required to complete five repetitions was recorded. The 6-MWT was conducted according to established guidelines [23], and the total walking distance was recorded. Body composition analysis via CT was conducted at baseline, after cycle 3, and at the end of chemotherapy. The cross-sectional areas of the psoas and skeletal muscles were measured at the level of the third lumbar vertebra (L3), a standardized anatomical landmark that correlates strongly with whole-body skeletal muscle mass and is widely used in sarcopenia research and oncology studies [24, 25]. Image analysis was performed using commercially available software (VINCENT, Fujifilm, Tokyo, Japan) with semi-automated segmentation based on predefined Hounsfield unit (HU) thresholds for muscle and adipose tissue. The skeletal muscle area was defined using a standard attenuation range of − 29 to + 150 HU. Volumes of the psoas muscle extending from the liver to the pubic symphysis, subcutaneous fat, and visceral fat were quantified with the Volume Analyzer application of the SYNAPSE VINCENT system. This system has been used in previous CT-based body composition studies [26, 27]. These CT scans were not performed specifically for this study but were part of routine clinical imaging conducted every three cycles during chemotherapy, allowing body composition analysis without additional examinations or patient burden. Adverse events were assessed at every study visit and graded according to the Common Terminology Criteria for Adverse Events, version 5.0. All adverse events were evaluated based on the clinical findings and laboratory data. Sample size The sample size was calculated based on detecting a mean difference of 4.5 with a standard deviation of 9.0, using estimates from previously published studies [28–31] and the minimal clinically important difference for the FACIT-F scale [32]. Assuming a two-sided alpha of 0.05 and a power of 70%—given the exploratory phase 2 design and feasibility considerations—and allowing for a 10% dropout rate, the final target sample size was set at 110 participants (55 per group). Statistical analyses All statistical analyses were conducted using JMP Pro version 17.0 (SAS institute Inc., Cary, NC, USA). Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range, IQR), as appropriate. Categorical variables were presented as frequencies and percentages. Between-group comparisons were performed using the independent samples t-test or the Mann–Whitney U test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables, depending on data distribution and sample size. For the primary endpoint, an analysis of covariance was conducted to compare mean values between the two groups, with performance status, cancer type, disease status (initial vs. recurrent), number of chemotherapy cycles, and study site as covariates. A two-sided test with a significance level of p < 0.05 was used to determine statistical significance. Moreover, a subgroup analysis of the primary outcome was conducted as a secondary analysis based on baseline physical activity levels prior to study initiation. Participants were classified into the following groups according to their weekly exercise habits:Group 1: No regular physical activityGroup 2: Low level of physical activity (i.e., < 150 min of moderate-to-vigorous physical activity and < 60 min of vigorous activity per week)Group 3: Sufficient level of physical activity (i.e., ≥ 150 min of moderate-to-vigorous physical activity and ≥ 60–120 min of vigorous activity per week) Group 1: No regular physical activity Group 1: No regular physical activity Group 2: Low level of physical activity (i.e., < 150 min of moderate-to-vigorous physical activity and < 60 min of vigorous activity per week) Group 2: Low level of physical activity (i.e., < 150 min of moderate-to-vigorous physical activity and < 60 min of vigorous activity per week) Group 3: Sufficient level of physical activity (i.e., ≥ 150 min of moderate-to-vigorous physical activity and ≥ 60–120 min of vigorous activity per week) Group 3: Sufficient level of physical activity (i.e., ≥ 150 min of moderate-to-vigorous physical activity and ≥ 60–120 min of vigorous activity per week) Secondary outcomes were analyzed using similar methods. All tests were two-tailed, and a p-value of < 0.05 was considered statistically significant. Statistical analyses were performed on the full analysis set, defined as all randomized patients who received at least one cycle of chemotherapy and underwent both baseline and at least one post-baseline assessment of the primary endpoint. Missing data were handled with a complete-case approach: for each analysis, including descriptive summaries and model-based comparisons, only participants with complete data for every required variable were included. Results A total of 110 patients were enrolled and randomized equally to the exercise and control groups (n = 55 each). In the exercise group, three patients withdrew consent, and one was deemed ineligible after enrollment, resulting in 51 evaluable cases (Figure S1). Baseline characteristics were well balanced between the two groups (Table 1). Adherence and activity data are summarized in Table 2. Table 1Patient characteristicsExercise group (n = 51)Control group (n = 55)Age (years), mean ± SD61.1 ± 10.0060.3 ± 11.80.732Performance status(ECOG), n(%) 051 (100)53 (96)0.496 102 (4)BMI (kg/m2), mean ± SD22.7 ± 0.6622.2 ± 0.630.548Cancer type, n(%) Ovarian21 (41)23 (42)0.872 Endometrial25 (49)25 (45) Cervical5 (10)7 (13)FIGO stage I34 (67)37 (67)0.659 II3 (6)5 (9) III8 (16)10 (18) IV6 (12)3 (6)Chemotherapy regimen TC45 (88)50 (91)0.755 TC་Bev6 (12)5 (9)Phase of chemotherapy Primary48 (94)50 (91)0.717 Recurrent3 (6)5 (9)Exercise habits 119 (37)30 (55)0.219 230 (59)23 (42) 32 (4)2 (4)Institution YCU42 (82)47 (85)0.792 YCUMC9 (18)8 (15)Baseline demographic and clinical characteristics of participants in the exercise and control groups. Values are presented as mean ± standard deviation (SD) or number (%). No significant differences were observed between groups across demographic or clinical variablesAbbreviations: ECOG Eastern Cooperative Oncology Group, BMI Body Mass Index, TC Paclitaxel + Carboplatin, Bev Bevacizumab, YCU Yokohama City University, YCUMC Yokohama City University Medical Center Table 1Patient characteristicsExercise group (n = 51)Control group (n = 55)Age (years), mean ± SD61.1 ± 10.0060.3 ± 11.80.732Performance status(ECOG), n(%) 051 (100)53 (96)0.496 102 (4)BMI (kg/m2), mean ± SD22.7 ± 0.6622.2 ± 0.630.548Cancer type, n(%) Ovarian21 (41)23 (42)0.872 Endometrial25 (49)25 (45) Cervical5 (10)7 (13)FIGO stage I34 (67)37 (67)0.659 II3 (6)5 (9) III8 (16)10 (18) IV6 (12)3 (6)Chemotherapy regimen TC45 (88)50 (91)0.755 TC་Bev6 (12)5 (9)Phase of chemotherapy Primary48 (94)50 (91)0.717 Recurrent3 (6)5 (9)Exercise habits 119 (37)30 (55)0.219 230 (59)23 (42) 32 (4)2 (4)Institution YCU42 (82)47 (85)0.792 YCUMC9 (18)8 (15)Baseline demographic and clinical characteristics of participants in the exercise and control groups. Values are presented as mean ± standard deviation (SD) or number (%). No significant differences were observed between groups across demographic or clinical variablesAbbreviations: ECOG Eastern Cooperative Oncology Group, BMI Body Mass Index, TC Paclitaxel + Carboplatin, Bev Bevacizumab, YCU Yokohama City University, YCUMC Yokohama City University Medical Center Patient characteristics Performance status (ECOG), n(%) Baseline demographic and clinical characteristics of participants in the exercise and control groups. Values are presented as mean ± standard deviation (SD) or number (%). No significant differences were observed between groups across demographic or clinical variables Abbreviations: ECOG Eastern Cooperative Oncology Group, BMI Body Mass Index, TC Paclitaxel + Carboplatin, Bev Bevacizumab, YCU Yokohama City University, YCUMC Yokohama City University Medical Center Table 2Exercise adherence and diary completionExercise group (n = 51)Control group (n = 55)Diary completion rate (%)83.1 ± 1.887.7 ± 1.7Patients meeting exercise prescription criteria† (n, %)39 (76%)39 (71%) *Mean walking time per week (min, Borg scale 11–13)184.3 ± 19.8178.3 ± 19.1Mean number of squats per day28.8 ± 4.2—†Exercise prescription criteria was defined as engaging in ≥ 90 min of walking at a Borg scale of 11–13 per week, across at least three sessions*Although no exercise prescription was given to the control group, 71% of participants exceeded the defined exercise criteria voluntarily— indicates data not applicable or not assessed Table 2Exercise adherence and diary completionExercise group (n = 51)Control group (n = 55)Diary completion rate (%)83.1 ± 1.887.7 ± 1.7Patients meeting exercise prescription criteria† (n, %)39 (76%)39 (71%) *Mean walking time per week (min, Borg scale 11–13)184.3 ± 19.8178.3 ± 19.1Mean number of squats per day28.8 ± 4.2—†Exercise prescription criteria was defined as engaging in ≥ 90 min of walking at a Borg scale of 11–13 per week, across at least three sessions*Although no exercise prescription was given to the control group, 71% of participants exceeded the defined exercise criteria voluntarily— indicates data not applicable or not assessed Exercise adherence and diary completion †Exercise prescription criteria was defined as engaging in ≥ 90 min of walking at a Borg scale of 11–13 per week, across at least three sessions *Although no exercise prescription was given to the control group, 71% of participants exceeded the defined exercise criteria voluntarily — indicates data not applicable or not assessed The mean diary completion rate was 83.1% (SD ± 1.8) in the exercise group and 87.7% (SD ± 1.7) in the control group. In the exercise group, 76% of participants met the predefined exercise prescription criteria, defined as engaging in ≥ 90 min of physical activity per week across at least three sessions. The mean weekly walking time (Borg scale 11–13) was 184.3 ± 19.8 min in the exercise group and 178.3 ± 19.1 min in the control group. The mean number of squats performed per day in exercise group was 28.8 ± 4.2. Notably, 71% of participants in the control group, despite receiving no formal exercise prescription, exceeded the same criteria. In both groups, all participants attended the scheduled visits as planned, except in cases where exercise was deemed medically inappropriate due to complications. No visits were cancelled at the patients’ request. When stratified by baseline exercise habits across both groups, 32 of 49 participants (65%) in Group 1 (no activity) met the exercise prescription criteria, compared to 46 of 57 (81%) in Groups 2–3 (some or sufficient activity). This difference suggested that participants with prior exercise habits were more likely to meet the criteria (odds ratio [OR]: 2.22, 95% confidence interval [CI]: 0.92–5.37, p = 0.07). Patient-related outcomes The FACIT and FACT scores indicate a better status with higher values and worse conditions with lower values, whereas higher scores in the PSQI reflect greater sleep disturbance. The mean change in the FACIT-F subscale score from baseline to post-treatment was 0.82 ± 7.8 in the exercise group and − 0.96 ± 9.2 in the control group, with no statistically significant difference between the groups (95% CI: -1.78 to 4.80, p = 0.36). Multivariate analysis showed no significant difference in changes in FACIT-F scores between the exercise and control groups (LS mean difference: 1.8, 95% CI: -1.3 to 4.9, p = 0.26). However, cancer type was significantly associated with changes in FACIT-F scores in the multivariable model (p < 0.01), with greater improvements observed in patients with ovarian cancer than in those with cervical (LS mean difference: 6.2, 95% CI: 0.9–11.6) or endometrial cancer (LS mean difference: 6.1, 95% CI: 2.7–9.5) [Table 3]. Table 3Multivariable linear regression analysis evaluating the association between cancer-related fatigue and covariates during chemotherapyVariableCategoryLS Mean Difference (95% CI)p-valueGroupExercise vs. Control1.8 (-1.3 to 4.9)0.26Performance status (ECOG)1–2 vs. 07.6 (-3.9 to 19.1)0.20Cancer typeOvarian vs. Cervical6.2 (0.9 to 11.6)< 0.01*Ovarian vs. Endometrial6.1 (2.7 to 9.5)Endometrial vs. Cervical0.14 (-5.1 to 5.4)Disease statusPrimary vs. Recurrent treatment4.2 (-2.1 to 10.6)0.19Number of chemotherapy cycles≥ 3 vs. <3-0.6 (-9.3 to 8.1)0.89Study siteYCU vs. YCUCC3.6 (-0.8 to 7.9)0.11“LS Mean Difference” represents the least squares mean difference and 95% confidence interval (CI)Abbreviations: ECOG Eastern Cooperative Oncology Group, YCU Yokohama City University, YCUCC Yokohama City University Medical Center*Statistically significant (p < 0.05) Table 3Multivariable linear regression analysis evaluating the association between cancer-related fatigue and covariates during chemotherapyVariableCategoryLS Mean Difference (95% CI)p-valueGroupExercise vs. Control1.8 (-1.3 to 4.9)0.26Performance status (ECOG)1–2 vs. 07.6 (-3.9 to 19.1)0.20Cancer typeOvarian vs. Cervical6.2 (0.9 to 11.6)< 0.01*Ovarian vs. Endometrial6.1 (2.7 to 9.5)Endometrial vs. Cervical0.14 (-5.1 to 5.4)Disease statusPrimary vs. Recurrent treatment4.2 (-2.1 to 10.6)0.19Number of chemotherapy cycles≥ 3 vs. <3-0.6 (-9.3 to 8.1)0.89Study siteYCU vs. YCUCC3.6 (-0.8 to 7.9)0.11“LS Mean Difference” represents the least squares mean difference and 95% confidence interval (CI)Abbreviations: ECOG Eastern Cooperative Oncology Group, YCU Yokohama City University, YCUCC Yokohama City University Medical Center*Statistically significant (p < 0.05) Multivariable linear regression analysis evaluating the association between cancer-related fatigue and covariates during chemotherapy “LS Mean Difference” represents the least squares mean difference and 95% confidence interval (CI) Abbreviations: ECOG Eastern Cooperative Oncology Group, YCU Yokohama City University, YCUCC Yokohama City University Medical Center *Statistically significant (p < 0.05) An exploratory subgroup analysis showed no significant differences in the FACIT-F scores between participants without baseline exercise habits and those with regular exercise habits (LS mean difference: 2.66, 95% CI: − 0.60 to 5.98). Figure 1 illustrates the overall profile of PROs using mean scores. The four subdomains of the FACT-G—physical well-being (PWB), social well-being (SWB), emotional well-being (EWB), and functional well-being (FWB)—were also assessed. Although the present study did not include a pre-specified analysis of the change from cycle 6 to post-treatment, visual inspection of the line graphs (Fig. 1) suggests a steeper recovery trajectory in the exercise group, particularly for the FACIT-F subscale score, FACT-G total scores, PWB, and EWB. This observation should be interpreted cautiously as it was not statistically tested. Fig. 1Longitudinal changes in PROs. Line graphs showing changes in mean FACIT-Fatigue, FACT-G, PWB, SWB, EWB, FWB, FACT/GOG-Ntx Subscale, and PSQI scores across time points in both groups. PROs = patient-reported outcomes, FACIT-F = Functional Assessment of Chronic Illness Therapy-Fatigue, FACT-G = Functional Assessment of Cancer Therapy-General, PWB = physical well-being, SWB = social well-being, EWB = emotional well-being, FWB = functional well-being, Ntx = neurotoxicity, PSQI = Pittsburgh Sleep Quality Index Longitudinal changes in PROs. Line graphs showing changes in mean FACIT-Fatigue, FACT-G, PWB, SWB, EWB, FWB, FACT/GOG-Ntx Subscale, and PSQI scores across time points in both groups. PROs = patient-reported outcomes, FACIT-F = Functional Assessment of Chronic Illness Therapy-Fatigue, FACT-G = Functional Assessment of Cancer Therapy-General, PWB = physical well-being, SWB = social well-being, EWB = emotional well-being, FWB = functional well-being, Ntx = neurotoxicity, PSQI = Pittsburgh Sleep Quality Index FACT/GOG-Ntx subscale scores deteriorated rapidly in both groups after chemotherapy initiation; however, a slight recovery trend was noted in the exercise group. PSQI scores were consistently lower in the exercise group throughout the treatment period; however, a statistically significant between-group difference was observed only after two cycles (estimated difference: 1.6 points, 95% CI: 0.2–2.9), with no significant differences at other time points. Physical function assessment Functional assessment results are summarized in Fig. 2. Fig. 2Changes in physical function before and after chemotherapy. Bar graphs illustrate pre- and post-treatment comparisons of physical function measures in the exercise and control groups. Evaluated parameters include quadriceps strength (right and left), quadriceps manual muscle testing (MMT) scores (right and left), handgrip strength (right and left), the five-times sit-to-stand test (5STS), and the 6-min walk distance. Light blue bars represent baseline values, and navy bars represent post-treatment values. Error bars represent mean ± standard deviation. For MMT scores, error bars were truncated at 5, the maximum possible value. For the 5STS, lower bounds of error bars were truncated at zero when the calculated value was negative Changes in physical function before and after chemotherapy. Bar graphs illustrate pre- and post-treatment comparisons of physical function measures in the exercise and control groups. Evaluated parameters include quadriceps strength (right and left), quadriceps manual muscle testing (MMT) scores (right and left), handgrip strength (right and left), the five-times sit-to-stand test (5STS), and the 6-min walk distance. Light blue bars represent baseline values, and navy bars represent post-treatment values. Error bars represent mean ± standard deviation. For MMT scores, error bars were truncated at 5, the maximum possible value. For the 5STS, lower bounds of error bars were truncated at zero when the calculated value was negative Handgrip strength showed a decreasing trend in both groups, and the time required for the 5STS test increased in both groups; however, no significant differences were observed between the groups. Quadriceps strength assessed by MMT remained unchanged in both groups. In contrast, measurements obtained using a handheld dynamometer indicated a trend toward preserved or slightly improved strength in the exercise group (right: +0.7 kgf; left: +1.1 kgf), whereas a decline was observed in the control group (right: −1.5 kgf; left: −0.8 kgf). The mean between-group difference in post-treatment change was 2.2 kgf (95% CI: 0.2 to 4.3; p = 0.035) for the right side and 1.9 kgf (95% CI: -0.2 to 4.0; p = 0.069) for the left side. In the 6MWT, the mean walking distance increased by 14 m in the exercise group and decreased by 5 m in the control group after treatment. The resulting between-group difference in change was 19 m but this was not statistically significant (95%CI: -13 to 51, p = 0.24). Body composition assessment Both groups exhibited minimal changes in body weight; however, reductions were observed in psoas muscle area, skeletal muscle area, psoas muscle volume, and subcutaneous fat volume, while visceral fat volume increased. None of these changes demonstrated significant differences between the groups (Table 4). Table 4Body composition assessmentMeasureGroupBaselineMean (SD)After 3 cyclesMean (SD)After treatmentMean (SD)ΔMean (SD)p-valueBody weight (kg)Exercise54.4 (11.0)-54.3 (10.5)-0.1 (3.0)0.44Control53.7 (11.0)-54.0 (11.4)0.3 (2.8)Psoas muscle area (cm2)Exercise9.3 (2.6)8.8 (2.4)9.0 (2.4)-0.2 (1.0)0.95Control8.8 (2.6)8.3 (2.5)8.6 (2.5)-0.3 (1.5)Skeletal muscle area (cm2)Exercise99.7 (18.9)94.7 (15.1)95.0 (16.4)-4.7 (8.3)0.38Control98.4 (16.7)94.2 (17.2)95.3 (17.6)-3.1 (9.9)Psoas muscle volume (cm3)Exercise185.3 (44.1)175.0 (45.7)179.7 (37.8)-5.6 (16.3)0.26Control187.9 (51.6)174.7 (42.3)176.5 (41.5)-11.4 (33.2)Subcutaneous fat volume (cm3)Exercise4141 (2225)3861 (2087)3858 (2176)-283 (1029)0.23Control3908 (2586)3703 (2171)3865 (2320)-43 (1007)Visceral fat volume (cm3)Exercise1461 (1106)1585 (1105)1622 (1156)161 (923)0.49Control1525 (1433)1559 (1301)1563 (1230)37 (905)Changes in body composition parameters before, during (after 3 cycles), and after chemotherapy in exercise and control groups. Values are shown as means and standard deviations (SD). Δ indicates the mean difference between baseline and post-treatment values. P-values reflect between-group comparisons of Δ values Table 4Body composition assessmentMeasureGroupBaselineMean (SD)After 3 cyclesMean (SD)After treatmentMean (SD)ΔMean (SD)p-valueBody weight (kg)Exercise54.4 (11.0)-54.3 (10.5)-0.1 (3.0)0.44Control53.7 (11.0)-54.0 (11.4)0.3 (2.8)Psoas muscle area (cm2)Exercise9.3 (2.6)8.8 (2.4)9.0 (2.4)-0.2 (1.0)0.95Control8.8 (2.6)8.3 (2.5)8.6 (2.5)-0.3 (1.5)Skeletal muscle area (cm2)Exercise99.7 (18.9)94.7 (15.1)95.0 (16.4)-4.7 (8.3)0.38Control98.4 (16.7)94.2 (17.2)95.3 (17.6)-3.1 (9.9)Psoas muscle volume (cm3)Exercise185.3 (44.1)175.0 (45.7)179.7 (37.8)-5.6 (16.3)0.26Control187.9 (51.6)174.7 (42.3)176.5 (41.5)-11.4 (33.2)Subcutaneous fat volume (cm3)Exercise4141 (2225)3861 (2087)3858 (2176)-283 (1029)0.23Control3908 (2586)3703 (2171)3865 (2320)-43 (1007)Visceral fat volume (cm3)Exercise1461 (1106)1585 (1105)1622 (1156)161 (923)0.49Control1525 (1433)1559 (1301)1563 (1230)37 (905)Changes in body composition parameters before, during (after 3 cycles), and after chemotherapy in exercise and control groups. Values are shown as means and standard deviations (SD). Δ indicates the mean difference between baseline and post-treatment values. P-values reflect between-group comparisons of Δ values Body composition assessment Changes in body composition parameters before, during (after 3 cycles), and after chemotherapy in exercise and control groups. Values are shown as means and standard deviations (SD). Δ indicates the mean difference between baseline and post-treatment values. P-values reflect between-group comparisons of Δ values Adverse events There were no significant differences between the groups in the mean number of chemotherapy cycles administered, the proportion of patients who experienced at least one treatment delay, the number of days delayed from the initially scheduled treatment completion date, or the proportion requiring dose reductions (Table 5). Table 5Chemotherapy administration and dose modificationsExercise group (n = 51)Control group (n = 55)p-valueCycles of chemotherapy5.7 ± 0.85.8 ± 0.70.45≥ 1 cycle chemotherapy delay, n (%)23 (45)23 (42)0.73Delayed treatment period (day)12140.45Cases of chemotherapy dose reductionone-level15 (29)16 (29)two-level2 (4)3 (5)1.00Summary of chemotherapy delivery and treatment modifications in the exercise and control groups. Values are expressed as mean ± SD or number (percentage). P-values indicate between-group comparisons Table 5Chemotherapy administration and dose modificationsExercise group (n = 51)Control group (n = 55)p-valueCycles of chemotherapy5.7 ± 0.85.8 ± 0.70.45≥ 1 cycle chemotherapy delay, n (%)23 (45)23 (42)0.73Delayed treatment period (day)12140.45Cases of chemotherapy dose reductionone-level15 (29)16 (29)two-level2 (4)3 (5)1.00Summary of chemotherapy delivery and treatment modifications in the exercise and control groups. Values are expressed as mean ± SD or number (percentage). P-values indicate between-group comparisons Chemotherapy administration and dose modifications Summary of chemotherapy delivery and treatment modifications in the exercise and control groups. Values are expressed as mean ± SD or number (percentage). P-values indicate between-group comparisons No significant differences were observed between the groups in the overall incidence of grade ≥ 3 hematologic toxicities or grade ≥ 2 non-hematologic toxicities (Tables 6 and 7). Table 6Grade ≥ 3 hematologic toxicitiesExercise group (n = 51)Control group (n = 55)p-valueneutropenia49 (96)51 (93)0.68anemia7 (14)12 (22)0.28thrombocytopenia5 (10)4 (7)0.74FN2 (4)2 (4)1.00Incidence of grade ≥ 3 hematologic adverse events in both groups. FN = febrile neutropenia Table 6Grade ≥ 3 hematologic toxicitiesExercise group (n = 51)Control group (n = 55)p-valueneutropenia49 (96)51 (93)0.68anemia7 (14)12 (22)0.28thrombocytopenia5 (10)4 (7)0.74FN2 (4)2 (4)1.00Incidence of grade ≥ 3 hematologic adverse events in both groups. FN = febrile neutropenia Grade ≥ 3 hematologic toxicities Incidence of grade ≥ 3 hematologic adverse events in both groups. FN = febrile neutropenia Table 7Grade ≥ 2 non-hematologic toxicitiesExercise group (n = 51)Control group (n = 55)p-valueNHT ≥Grade 2, n (%)48 (94)48 (87)0.32nausea3 (6)15 (27)0.004*anorexia10 (20)12 (22)0.78peripheral neuropathy43 (84)42 (76)0.30arthralgia14 (27)22 (40)0.17myalgia5 (10)14 (25)0.044*constipation12 (24)18 (33)0.29diarrhea4 (8)6 (11)0.59fatigue8 (16)16 (29)0.099skin rash6 (12)2 (4)0.10hypertension1 (2)3 (5)0.35Frequency of grade ≥ 2 non-hematologic toxicities during chemotherapy. Non-hematologic toxicity (NHT) categories include nausea, fatigue, neuropathy, and others*Statistically significant difference between the groups (p < 0.05) Table 7Grade ≥ 2 non-hematologic toxicitiesExercise group (n = 51)Control group (n = 55)p-valueNHT ≥Grade 2, n (%)48 (94)48 (87)0.32nausea3 (6)15 (27)0.004*anorexia10 (20)12 (22)0.78peripheral neuropathy43 (84)42 (76)0.30arthralgia14 (27)22 (40)0.17myalgia5 (10)14 (25)0.044*constipation12 (24)18 (33)0.29diarrhea4 (8)6 (11)0.59fatigue8 (16)16 (29)0.099skin rash6 (12)2 (4)0.10hypertension1 (2)3 (5)0.35Frequency of grade ≥ 2 non-hematologic toxicities during chemotherapy. Non-hematologic toxicity (NHT) categories include nausea, fatigue, neuropathy, and others*Statistically significant difference between the groups (p < 0.05) Grade ≥ 2 non-hematologic toxicities Frequency of grade ≥ 2 non-hematologic toxicities during chemotherapy. Non-hematologic toxicity (NHT) categories include nausea, fatigue, neuropathy, and others *Statistically significant difference between the groups (p < 0.05) However, among the grade ≥ 2 non-hematologic toxicities, the incidence of nausea and myalgia was significantly lower in the exercise group. Specifically, nausea occurred in three patients (6%) in the exercise group and 15 patients (27%) in the control group (OR: 0.17, 95% CI: 0.05–0.62, p = 0.004), while myalgia was reported in five (10%) and 14 patients (25%), respectively (OR: 0.32, 95% CI: 0.11–0.96, p = 0.044). In addition, no adverse events related to the exercise intervention, such as knee pain due to improper squat execution, were noted. Discussion In this prospective RCT evaluating the effects of home-based exercise intervention during chemotherapy for gynecologic cancers, no statistically significant difference was observed between the exercise and control groups in the primary outcome, namely changes in the FACIT-F score. Although the study was designed to detect a clinically meaningful difference, this threshold was not met, and the primary hypothesis was not supported. Several factors likely contributed to the lack of between-group differences. First, participants in the control group engaged in unexpectedly high levels of voluntary physical activity. Despite the absence of structured guidance, 71% met the exercise criteria. This contamination likely diminished the intervention effect. This may reflect increased exercise awareness at trial enrollment or behavioral activation prompted by the use of daily exercise diaries in both groups. Supporting this, Milzer et al. reported that journaling enhances self-management behaviors in patients with cancer, which may have similarly influenced behavior in this study [33]. Second, the limited level of supervision in this study may have attenuated the effect of the intervention. Although participants in the exercise group attended the rehabilitation department every 3 weeks, the program was largely home-based and relied on self-directed exercises. This level of supervision may have been insufficient to ensure optimal exercise intensity, progression, and fidelity to the prescribed programs. In particular, adherence was assessed from self-reported exercise diaries, and it remains unclear whether the prescribed intensity (Borg scale 11–13) was consistently achieved in real-world settings. Although adherence in this study was relatively high compared with prior reports, the lack of close monitoring may have led to variability in the actual exercise dose and quality. Supervised exercise programs tend to produce better outcomes than unsupervised, home-based regimens [34, 35]. Moreover, studies conducted in other countries have demonstrated improved adherence through remotely delivered support from nurses or certified cancer exercise trainers, often incorporating CBT [10, 36]. Therefore, more intensive supervision, such as frequent feedback or real-time monitoring, may have further enhanced both adherence and intervention fidelity in our study. However, implementing such resource-intensive supervision in routine clinical practice remains challenging in Japan due to workforce and systemic constraints. Recently, digital health tools, such as smartphone-based symptom monitoring applications, have shown benefits in improving QoL and managing side effects among patients with cancer [37]. These technologies may serve as platforms for delivering exercise support in future interventions. Third, the study period included a phase overlapping with the COVID-19 pandemic, and 69% of the participants were enrolled between July 2020 and May 2023, during which government-imposed limitations on daily activities were in place. Findings from a retrospective study conducted by our research group during the pandemic suggested a tendency toward higher levels of anxiety and depression among patients undergoing cancer treatment than in pre-pandemic reports [38]. Therefore, the potential impact of the COVID-19 pandemic on participants’ psychological status and perceptions of fatigue cannot be ruled out. Although physical activity levels remained relatively high in both groups, pandemic-related psychological stress may have influenced fatigue perception, independent of exercise volume. Despite no significant effect on the primary outcome, several secondary outcomes showed changes, suggesting potential benefits in the exercise group. In terms of QOL, changes toward improvement were observed in physical and emotional well-being (PWB and EWB subdomains of FACT-G) as well as in the total FACT-G score, particularly between the end of chemotherapy (cycle 6) and the post-treatment assessment, although these findings were based on visual inspection without formal statistical testing. Additionally, quadriceps strength was better preserved or improved, suggesting a protective effect on lower limb function. The preservation of lower limb muscle strength, including quadriceps strength, may have clinical relevance. Reduced lower limb strength is associated with poorer survival in patients with lung and colorectal cancers [39, 40]. Therefore, the observed trend toward preserved quadriceps strength in the present study may suggest not only a protective effect on physical function but also a potential contribution to long-term health outcomes. Furthermore, the incidence of grade ≥ 2 nausea and myalgia was significantly lower in the exercise group. Although similar findings have been reported in breast cancer [41], the findings of our study, reported for the first time in gynecologic cancer, indicate a potential role of exercise in alleviating fatigue and managing chemotherapy-related side effects. Two additional findings merit particular attention. First, heterogeneity in patient characteristics appeared to influence both the outcomes and adherence. Multivariable analysis revealed significant improvement in FACIT-F scores among patients with ovarian cancer. As this study did not include further stratified analyses by cancer type, the observed trend remains hypothetical. However, patients with ovarian cancer typically present with a higher baseline symptom burden—including fatigue, abdominal symptoms, and dyspnea—partly due to peritoneal dissemination and ascites [42]; additionally, a large multicancer study demonstrated that baseline health-related QoL scores were lower in patients with ovarian cancer than in those with other malignancies [43]. These features may collectively increase susceptibility to cancer-related fatigue in patients with ovarian cancer, suggesting that ovarian cancer may represent a hypothesis-generating subgroup for future exercise trials. In addition, an exploratory subgroup analysis showed that participants with a prior exercise habit had higher adherence rates to the intervention, a trend that aligns with the findings of previous reviews identifying extensive exercise history as a predictor of better adherence to exercise interventions in cancer populations [44]. This suggests that tailoring strategies based on baseline physical activity levels may be key to improving both adherence and outcomes. For example, individuals with an established exercise routine may benefit from low-touch interventions such as diaries or occasional encouragement, while those without such habits may require more structured or supervised programs. These insights may inform the design of future interventions. Another important finding was that CT-based body composition analysis revealed reductions in skeletal muscle mass and increases in visceral fat in both groups, despite the preservation of lower limb strength in the exercise group. These findings suggest that exercise alone is insufficient to mitigate treatment-related changes in body composition, underscoring the importance of incorporating nutritional support in future studies. The ongoing BENITA trial, which integrates app-based exercise and nutritional support and includes outcomes such as 6-min walk distance, nutritional status, and body composition, is expected to provide valuable insights for developing more effective and individualized supportive care strategies [45]. Strengths and limitations This study has several notable strengths. To our knowledge, it is one of the largest RCTs to date evaluating exercise interventions during active chemotherapy in patients with gynecologic cancers. The study employed a comprehensive evaluation framework, including PROs, objective physical performance measures, and CT-based body composition analysis, allowing multidimensional assessment of the intervention’s impact. Furthermore, the pragmatic, home-based design enhances the real-world applicability of our findings and supports the feasibility of incorporating similar low-cost, scalable interventions into routine cancer care. However, several limitations should be acknowledged. First, due to the nature of the intervention, blinding was not feasible, and participants may have had preconceived expectations of benefit. Second, ethical considerations precluded restricting voluntary exercise in the control group, which may have influenced outcomes. Third, all participants were Japanese, which may limit generalizability to more diverse populations. Fourth, age-related differences in the effectiveness of the interventions were not evaluated in this study. Finally, the social context of the COVID-19 pandemic may have influenced participant behavior, exercise compliance, and the feasibility of intervention delivery, which should be considered when interpreting the results. Conclusions This RCT demonstrated that a home-based exercise program was safe and feasible during chemotherapy for gynecologic cancer. The primary endpoint, change in the FACIT-F fatigue score, did not differ significantly between the exercise and control groups, possibly due to contamination in the control group. Nonetheless, some between-group differences were observed in physical function, quality of life, and non-hematologic toxicities, and the high adherence rates observed in both groups indicate that exercise during chemotherapy is achievable and well accepted by patients, reflecting patients’ strong intrinsic motivation to maintain physical activity. These findings suggest that exercise can serve as a supportive approach to mitigating treatment-related side effects and preserving functional capacity during chemotherapy. Future studies should refine exercise protocols through digital monitoring and combined nutritional strategies to enhance efficacy while identifying patient subgroups most likely to benefit during active cancer treatment. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.