Effects of activity pacing–related interventions on cancer-related fatigue and classification of behavior change techniques: A systematic review and meta-analysis Authors: Yoichi Shimizu, Katsunori Tsuji, Akiko Hanai, Mai Hosokawa Published: 2026 ABSTRACT AbstractPurposeAlthough physical activity is recommended for managing cancer-related fatigue (CRF), activity pacing (AP) and energy conservation are also implemented to balance activity and rest. This systematic review assessed the effectiveness of AP-related interventions and behavior change techniques (BCTs) employed in AP-related interventions for CRF management.MethodsWe searched PubMed, CINAHL, CENTRAL, and Ichu-shi databases for articles published up to February 28, 2026. After assessing the risk of bias, we performed meta-analyses for quantitative synthesis. We also identified and classified BCTs using data from the Behavior Change Technique Taxonomy version 1.ResultsAmong the 1257 identified studies, 10 (7 randomized controlled trials [RCTs] and 3 non-RCTs) met the inclusion criteria. The effect size was small but not significant (standardized mean differences [SMD] = − 0.36; 95% confidence intervals [CI]: − 0.73 to 0.00; 5 studies; 617 participants), with substantial heterogeneity (I2 = 66%). Compared with controls, the intervention showed a moderate, statistically significant effect in the subgroup of patients undergoing chemotherapy (SMD = − 0.58; 95% CI: − 0.87 to − 0.29; three studies; 197 participants with breast cancer). The primary BCTs used in AP-related interventions were “goal setting,” “social support,” and “graded tasks.”ConclusionAP-related interventions may provide some benefit in improving CRF, particularly during chemotherapy. However, the certainty of the current evidence is extremely low. Although these interventions represent useful approaches, further RCTs are needed, and digitization of AP-related intervention programs incorporating effective BCTs may be beneficial.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00520-026-10754-3.Keywords: Activity pacing, Cancer-related fatigue, Neoplasms, Supportive care, Cancer Background Cancer-related fatigue (CRF) is a painful, persistent subjective sensation of physical, emotional, and/or cognitive fatigue or exhaustion associated with cancer or cancer treatment that is experienced by approximately 65% of patients with cancer. Additionally, CRF is one of the most distressing symptoms of the disease [1–3]. In addition to pharmacological treatments, nonpharmacological alternatives are implemented to manage CRF [3], including exercise therapy and psychological interventions, which are more effective than pharmacological treatments [4]. The European Society of Medical Oncology guidelines for CRF management recommend regular exercise or physical activity in the absence of cachexia [2]. Aerobic exercise (walking, swimming, and light jogging), light muscle training, and gentle stretching improve muscle strength and physical function as well as reduce fatigue. To determine an appropriate activity level based on a patient’s abilities and treatment status, consultation with a health or exercise professional is important. However, healthcare providers do not adequately guide these patients regarding exercise [5], and adherence to guideline-recommended regular exercise is low [6, 7]. Although exercise therapy is recommended for CRF, activity pacing (AP) and energy conservation techniques are also implemented clinically. AP divides daily activities into smaller, more manageable portions in a manner that minimizes symptom exacerbation, thereby allowing gradual, progressive increases in physical activity [8]. The goals of AP are to separate the symptom experience from the activity experience, prevent overexertion, attenuate fluctuations in physical activity patterns, and prevent the adverse consequences associated with fatigue-induced inactivity. The key components of AP are conserving energy for valued activities (i.e., playing with children or engaging in recreational activities) and setting graded activity quotas that allow gradual increases in the ability to perform activities [9]. Conversely, energy conservation is the deliberate, planned management of an individual’s energy resources to prevent their depletion [10]. The objective of energy conservation is to balance rest and activity during periods of high fatigue, enabling the maintenance of valued activities and goals [10]. Although these interventions have often been described using different terminologies, several conceptual similarities have been noted. Both activity pacing and energy conservation emphasize the regulation of daily activities and efficient use of available energy to reduce fatigue and maintain participation in daily life. Therefore, previous clinical trials have implemented related strategies under various labels, including energy conservation training [11], energy conservation and activity management [10], and AP programs [12]. However, evidence supporting the efficacy of these interventions remains fragmented because of variations in terminology and intervention descriptions across studies. Consequently, the overall effectiveness of behavioral energy management strategies for CRF remains unclear. Synthesizing these interventions (hereafter termed AP-related interventions) within a unified analytical framework may provide a clearer understanding of their potential benefits. Moreover, CRF is mechanistically and clinically distinct from chronic fatigue syndrome (CFS). CFS is a multisystem condition characterized by postexertional malaise, autonomic dysfunction, and chronic fatigue unrelated to cancer therapies [13]. In contrast, CRF arises from cancer biology, treatment side effects, inflammation, anemia, and psychological distress [3]. Therefore, findings from AP-related interventions in CFS cannot be directly generalized to patients with cancer. Although a 2023 systematic review examined the effectiveness of AP for CFS [14], no systematic synthesis has focused specifically on AP interventions for CRF. Considering the differences in symptom mechanisms, treatment contexts, and functional limitations, a dedicated review of AP for CRF is warranted. Thus, this review aimed to determine the effectiveness of AP-related interventions compared with usual care in improving CRF. The secondary objective was to synthesize the behavior change techniques (BCTs) used in AP-related interventions for CRF management. Methods Protocol and registration The protocol for this review was preregistered on PROSPERO (registration number: CRD42020169986) and adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist [15]. Eligibility criteria The inclusion criteria for this study were as follows: (1) randomized controlled trials (RCTs) or nonrandomized interventional studies; (2) adult participants with cancer and CRF; and (3) studies reporting the effectiveness of AP or energy conservation techniques on CRF as a primary outcome. We excluded protocol papers, case reports, and RCTs involving pediatric populations. Information sources and search strategy A comprehensive search of the PubMed, CINAHL, CENTRAL, and Ichu-shi databases was conducted to identify studies published up to February 28, 2026. The search strategy included terms related to “activity pacing” or “energy conservation techniques” such as (Conservation of Energy Resources) OR (Energy Conservation Technique) OR (Energy Conservation Therapy) OR (Activity pacing) and terms related to “cancer-related fatigue” such as [{fatigue} OR {tired*} OR {exhaustion}] OR [{lethargy} OR {asthenia}]. All search terms are presented in Supplemental Table 1. All studies that evaluated AP-related interventions in patients with cancer were considered. Study selection and data extraction Studies that met the eligibility criteria were assessed using a two-step process. The retrieved references were imported into the Rayyan web application (https://rayyan.ai/), a tool designed for systematic reviews, to facilitate the screening process [16]. First, two authors (MH and YS) independently reviewed the titles and abstracts of the identified studies and then reviewed the full-text articles. All disagreements were discussed between the reviewers, and two other authors (KT and AH) were consulted to resolve the remaining disagreements through consensus. The authors independently extracted data from each included study. The extracted data included the following: identification information (publication year and first author), study information (purpose, study design, and study duration), general information (sample size, country, and study design), participant characteristics (cancer type, age, and sex), intervention details (intervention type, intervention frequency, and duration), outcomes (data and time points for each measurement), and fatigue evaluation results. Outcome measures The primary outcome was change in CRF assessed using validated fatigue measurement scales, such as the Revised Piper Fatigue Scale and Brief Fatigue Inventory. The secondary outcomes were changes in physical activity levels, quality of life, and adherence to AP-related interventions. Risk of bias assessment For RCTs, two reviewers (MH and YS) independently evaluated the risk of bias using the Cochrane Risk of Bias 2 tool (ROB-2) [17, 18], which comprises the following five domains: randomization process, deviations from the intended interventions, missing outcome data, outcome measurement, and selection of the reported result. Each domain was assigned a risk level, namely, “low risk,” “some concerns,” or “high risk.” Then, two researchers (AH and KT) independently evaluated the certainty of the evidence using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) the Evidence to Decision framework via GRADEpro GDT (McMaster University and Evidence Prime, 2023), and all researchers reached consensus on the results. For nonrandomized interventional studies, two reviewers (MH and KT) independently evaluated the risk of bias using the Risk Of Bias In Non-randomized Studies-of Interventions, Version 2 (ROBINS-I V2) tool [19, 20], which comprises the following seven domains: confounding, intervention classification, participant selection into the study (or into the analysis), deviations from intended interventions, missing data, outcome measurement, and selection of the reported result. Each domain was assigned a risk level, namely, “low risk,” “moderate risk,” “serious risk,” or “critical risk.” Statistical analysis for quantitative synthesis We summarized all data extracted from each study and conducted a meta-analysis of the RCTs using the Review Manager 5.4 (RevMan 5.4) software. Because different fatigue scales were used across studies, we calculated the effect sizes using standardized mean differences (SMD) with 95% confidence intervals (CIs). We employed a random-effects model to account for clinical heterogeneity among studies and applied the principles of intention-to-treat principle. The following Cohen’s effect sizes were used: 0.2, small effect; 0.5, moderate effect; and 0.8, large effect [21]. Furthermore, standard deviations (SDs) for endpoint scores were used when SDs for change scores were unavailable. We assessed heterogeneity using the I2 statistic, interpreted as follows: 0%–40% indicates insignificant heterogeneity, 30%–60% indicates moderate heterogeneity, 50%–90% indicates substantial heterogeneity, and 75%–100% indicates considerable heterogeneity [22]. Publication bias assessment using funnel plots [23] and statistical tests (e.g., Egger’s test) was planned when at least 10 studies were available for a given outcome. Because fewer than 10 studies were included in the meta-analysis, formal assessment of publication bias was not conducted, in accordance with methodological recommendations. Finally, subgroup analyses were performed for the group undergoing chemotherapy and for the group undergoing or having completed radiotherapy. Sensitivity analyses were performed using a leave-one-out approach, in which each study was sequentially excluded to evaluate its influence on the pooled effect estimates. GPT-assisted BCT extraction and coding We used a custom GPT application within ChatGPT (“Make BCTT list by research paper,” developed by AH) provided by OpenAI. The model deployed at runtime was GPT-4o, and coding was conducted in September 2025. The model was not fine-tuned for this task. The custom GPT was configured through structured task instructions and two reference documents uploaded as knowledge sources: the Behaviour Change Technique Taxonomy v1 (BCTTv1) manual and the foundational publication by Michie et al. [24]. A standardized prompt was applied to all papers to instruct the model to identify BCTs using BCTTv1 codes and labels and to cite supporting text from the source article with page and line references. The complete standardized prompt is provided verbatim in Supplementary Material 1. GPT-generated outputs were used to support decision-making. For each included study, all four authors reviewed the GPT-identified BCTs alongside the original text and reached consensus on the final BCT codes and labels included in the dataset. GPT was used solely as a decision-support tool, and all final coding decisions were made by the authors. Because coding was finalized through consensus rather than independent parallel coding, inter-rater agreement statistics were not calculated. Results Study selection Overall, 1257 studies were identified, with 10 (7 RCTs [11, 22–27] and 3 nonrandomized studies [28–30]) meeting the inclusion criteria. The study selection process is presented in Fig. 1. Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram of the study selection process for the systematic review Study characteristics The characteristics of the seven RCTs [10–12, 25–28] and three nonrandomized studies [29–31] are presented in Table 1. Six studies focused on patients undergoing chemotherapy, with all interventions provided during treatment [11, 12, 26–28, 30]. One study focused on patients who had completed radiotherapy [25]. Three studies focused on patients undergoing chemotherapy, radiotherapy, or chemoradiotherapy [10, 29, 31]. No study focused on end-of-life patients. Although the rating scales varied across the studies, the Revised Piper Fatigue Scale and Brief Fatigue Inventory were the most used. The rating scales are presented in Table 2. Table 1 Study characteristicsItemRCT (n = 7)Non-RCT (n = 3)n(%)n(%)Cancer typeBreast5(71.4)1(33.3)Cervical1(33.3)Any cancer2(28.6)1(33.3)StageI–II1(14.3)I–III3(42.9)1(33.3)Any stage2(28.6)1(33.3)Unknown1(14.3)1(33.3)Treatment typeChemotherapy5(71.4)1(33.3)Radiotherapy, chemotherapy, or concurrent chemoradiotherapy1(14.3)2(66.7)After radiotherapy1(14.3)CountryUnited States2(28.6)1(33.3)Ethiopia1(14.3)1(33.3)Iran2(28.6)Thailand2(28.6)China1(33.3)Number of eligible participants > 1011(14.3)2(66.7) < 1006(85.7)1(33.3)RCT, Randomized controlled trial Study characteristics RCT, Randomized controlled trial Table 2 Outcome measures RCT (n = 7)Non-RCT (n = 3)n(%)n%Revised Piper Fatigue Scale2(28.6)1(33.3)Brief Fatigue Inventory1(14.3)1(33.3)Cancer Fatigue Scale2(28.6)Profile of Mood States1(14.3)1(33.3)Numeric Rating Scale1(14.3)Fatigue-VAS1(14.3)Piper Fatigue Scale1(14.3)Schwartz Fatigue Scale1(14.3)General Fatigue Scale1(14.3)The total exceeds 100% because some papers used multiple measuresRCT, Randomized controlled trial; VAS, Visual analog scale Outcome measures The total exceeds 100% because some papers used multiple measures RCT, Randomized controlled trial; VAS, Visual analog scale A summary of the characteristics of each study is presented in Tables 3 and 4. Most studies had a relatively small number of participants and focused on patients with breast cancer. No adverse events were reported. Table 3 Characteristics of eligible RCTs First author, year, and countryParticipantsCancer treatmentSummary of the interventionOutcome measureSummary of results Barsevick 2004, United States Sample size: IG, 200; CG, 196Age: 56.3 ± 12.5 yearsSex: Male, 59 (15%); female, 337 (85%)Cancer type: Breast, 282 (71%); lung, 63 (16%); lymphoma, 24 (6%); colorectal, 16 (4%); and genital, 11 (3%)Cancer stage: 0, 25 (6%); I, 120 (30%); II, 128 (32%); III, 75 (19%); IV, 35 (9%); and Unavailable, 13 (4%)Timing: At the beginning of treatmentTreatment: Chemotherapy only, 184 (47%); Radiotherapy only, 176 (44%); and Concurrent therapy, 35 (9%)Regimen: UnknownIntervention details: IG (ECAM): Goal setting, prioritization, and planning based on the Common Sense Model tenetsS1: Received information on CRF and learned energy conservation skillsS2: Assisted in developing an energy conservation plan to manage valued activities and to minimize fatigue interferenceS3: Evaluated and revised the planCG: Received standard nutrition informationResearch period: 5 weeks;Intervention frequency: Weekly (telephone sessions), intervention sessions (30 min each for Sessions 1 and 2 and 15 min for Session 3)Measurements: POMS, SCFS, and GFSTiming of rating: Patients receiving chemotherapy: pretreatment, 48 h after second chemotherapy, and 48 h after third chemotherapyPatients receiving radiotherapy: pretreatment, last week of radiotherapy, and 1 month after radiotherapy completionThe significant group-by-time interaction revealed that fatigue, as measured by the GFS, POMS-F, and SCFS_Physical scales, was significantly lower over time in the ECAM group than in the CG (P = 0.01, P = 0.05, and P = 0.05, respectively) Yuen 2006, United States Sample size: IG, 6; CG, 6Age: 55.4 ± 9.3 years;Sex: Male, 7 (58%); Female, 5 (42%)Cancer type: UnknownCancer stage: UnknownTiming: After radiation therapyTreatment: RadiotherapyIntervention details: IG (Energy conservation): 1–2 h Face-to-face training and 3 telephone follow-ups; CG (usual care): Standard careResearch period: From 2 to 4 weeks after therapyIntervention frequency: One in-person training and weekly telephone follow-ups for 3 weeksMeasurements: R-PFSTiming of rating: Preassessment within 2 weeks after therapy and postassessment after 3 telephone follow-upsThe pretraining PFS scores significantly differed between the energy conservation training and usual care groups Sadeghi 2016, Iran Sample size: IG, 69; CG, 66Age: IG, 55.7 ± 10.2 years; CG, 55.6 ± 13.7 yearsSex: Male, 0 (0%); female, 135 (100%)Cancer type: Breast, 135 (100%)Cancer stage: IG: I, 13 (19%); Ⅱ, 28 (43%); and III, 25 (38%); CG: I, 13 (20%); II, 31 (44%); and III, 25 (36%)Timing: At the start of treatmentTreatment: ChemotherapyRegimen: UnknownIntervention details: IG (Energy conservation training): Goal setting, activity prioritization, journaling, development of energy management skills education on work–life balance, discussion on resource use, and practice and homework to monitor their fatigue and activityCG: Standard care only and patients were placed on a waitlistResearch period: 5 weeksIntervention frequency: 5 weekly 90-min group sessionsMeasurements: CFSTiming of rating: Preintervention, postintervention, and 8-week follow-upIG: CRF was reduced from pre- to postintervention and continued to decrease over the 8-week follow-up period (F = 69.8, P < 0.001)CG: The CFS levels showed no statistically significant increases over time Ninsen 2021, Thailand Sample size: IG, 37; CG, 37Age: IG, 50.1 ± 8.6 years;CG, 52.8 ± 10.1 yearsSex: Male, 0 (0%); female, 74 (100%)Type of cancer: Breast, 74 (100%)Cancer stage: IG: I, 12 (32%); II, 17(46%); and III, 8 (21.6%); CG: I, 14 (38%); II, 17 (46%); and III, 6 (16%)Timing: During chemotherapy;Treatment: Chemotherapy;Regimen: AC regimen onlyDetails of the intervention: IG (energy conservation program): Day 1: 30 min face-to-face health education in symptom management; days 2 and 7: 20 min telephone call assessment of self-care activity and monitoring CG: Energy conservation program in the study focusedResearch periods: 2 weeksFrequency of intervention: Days 1, 2, and 7 on AC therapyMeasurements: NRSTiming of rating: Days 1 (baseline), 2, 7, and 14Postintervention period:The median fatigue score in the IG improved significantly at the postintervention period compared with that in the CG. Cycle 1: The score was also significantly lower in the IG than that in the CG on day 14 (P < 0.01)Cycle 2: Fatigue scores on days 7 and 14 in the IG were significantly lower than those in the CG group (P = 0.02 and 0.01, respectively) Cycle 3: Only the fatigue score on day 7 in the IG was significantly lower than that in the CG (P = 0.04)Cycle 4: Fatigue scores on days 1, 2, 7, and 14 in the IG were significantly lower than those in the CG (P < 0.001, 0.04, 0.01, and 0.01, respectively) Sharbafchi 2023, Iran Sample size: IG, 25; CG, 25;Age: IG, 45.9 ± 7.6 years; CG, 46.7 ± 1.7 yearsSex: Male, 0 (0%); female, 50 (100%)Cancer type: Breast, 50 (100%)Cancer stage: IG: I, 7 (32%); II, 15 (68%). CG: I: 8 (38%); II, 13 (62%)Timing: During the third-to-fourth cycle of chemotherapy;Treatment: ChemotherapyRegimen: UnknownIntervention details: IG (energy conservation and concentrative movement therapy): Session-based energy management and self-assessment, fatigue assessment, general education on fatigue, and practice in energy conservation CG: Three 90-min weekly educational sessions on self-management during chemotherapyResearch periods: 8 weeks;Frequency of intervention: Weekly, three 90-min sessionsMeasurements: F-VAS, CFS, and R-PFSTiming of rating: Pretreatment, post-treatment, 1 week after treatment, and 4 weeks after treatmentIn all time points (after the last session, 1 week, and 4 weeks), F-VAS, CFS, and PFS scores significantly differed between the IG and CG (P < 0.001) All three CFS subscales (physical, affective, and cognitive) significantly decreased in the IG compared with those in the CG (P < 0.001)Naraphong 2023, ThailandSample size: IG, 10; CG, 9Age: IG, 43.5 ± 9.1 years;CG, 48.8 ± 6.5 yearsSex: Male, 0 (0%); female, 19 (100%)Cancer type: Breast, 19 (100%);Cancer stage: IG: I, 1 (10%); II, 8 (80%); and III, 1 (10%). CG: I: 1 (11%); II, 5 (56%); and III, 3 (33%)Timing: During chemotherapyTreatment: ChemotherapyRegimen: UnknownIntervention details: IG (ECAM): Prioritization, planning, stress management, and journaling. CG: Provision of general health information onlyResearch period: 12 weeks;Intervention frequency: WeeklyMeasurements: R-PFS;Timing of rating: Preintervention and 12 weeksFatigue scores at 12 weeks were lower in the IG (X = 1.41, SD = 1.35) than in the CG (X = 3.32, SD = 1.42) The fatigue score from baseline to week 12 showed no significant difference within each groupIndependent t-test results revealed that the scores for fatigue intensity at week 12 significantly differed between the IG and CG (t[17] = − 3.006, P = 0.008) The effect size for fatigue change was determined using Hedges’ g (g = − 1.38; 95% CI [− 2.35, − 0.41]) Getu 2023, Ethiopia Sample size: IG, 30; CG, 28Age: IG, 40.2 ± 10.9 years;CG, 42.5 ± 12.3 yearsSex: Male, 0 (0%) and Female, 58 (100%)Cancer type: Breast, 58 (100%)Cancer stage: IG: I, 3 (10%); II, 15 (50%); III, 11 (37%); and IV, 1 (3.3%). CG: I: 2 (7%); II, 10 (36%); III, 13 (46%); and IV, 3 (11%)Timing: During chemotherapyTreatment: ChemotherapyRegimen: UnknownIntervention details: IG (cognitive behavioral therapy with integrated activity pacing): SMART goal setting, relaxation techniques, and cognitive restructuringCG: Standard care onlyResearch period: 7 weeks;Frequency of intervention: 3 face-to-face sessions and 4 phone sessionsMeasurements: BFITiming of rating: Preintervention, postintervention (end of the intervention), and 3-month follow-upRM-ANCOVA showed a significant reduction in fatigue in the IG compared with that in the CG from baseline to the end of intervention and at the 3-month follow-up, based on the group × time interaction (P < 0.001, np2 = 0.596)AC:, BFI, Brief Fatigue Inventory; CFS, Cancer Fatigue Scale; CG, control group; CRF, cancer-related fatigue; ECAM, energy conservation and activity management; F-VAS, Fatigue Visual Analog Scale; GFS, General Fatigue Scale; IG, intervention group; NRS, Numeric Rating Scale; POMS, Profile of Mood States; RCTs, randomized controlled trials; RM-ANCOVA, repeated measures analysis of covariance; R-PFS, Revised Piper Fatigue Scale; SCFS, Schwartz Cancer Fatigue Scale; SMART goals, Specific, Measurable, Achievable, Relevant, and Time-bound goals Characteristics of eligible RCTs Sample size: IG, 200; CG, 196 Age: 56.3 ± 12.5 years Sex: Male, 59 (15%); female, 337 (85%) Cancer type: Breast, 282 (71%); lung, 63 (16%); lymphoma, 24 (6%); colorectal, 16 (4%); and genital, 11 (3%) Cancer stage: 0, 25 (6%); I, 120 (30%); II, 128 (32%); III, 75 (19%); IV, 35 (9%); and Unavailable, 13 (4%) Timing: At the beginning of treatment Treatment: Chemotherapy only, 184 (47%); Radiotherapy only, 176 (44%); and Concurrent therapy, 35 (9%) Regimen: Unknown Intervention details: IG (ECAM): Goal setting, prioritization, and planning based on the Common Sense Model tenets S1: Received information on CRF and learned energy conservation skills S2: Assisted in developing an energy conservation plan to manage valued activities and to minimize fatigue interference S3: Evaluated and revised the plan CG: Received standard nutrition information Research period: 5 weeks; Intervention frequency: Weekly (telephone sessions), intervention sessions (30 min each for Sessions 1 and 2 and 15 min for Session 3) Measurements: POMS, SCFS, and GFS Timing of rating: Patients receiving chemotherapy: pretreatment, 48 h after second chemotherapy, and 48 h after third chemotherapy Patients receiving radiotherapy: pretreatment, last week of radiotherapy, and 1 month after radiotherapy completion Sample size: IG, 6; CG, 6 Age: 55.4 ± 9.3 years ; Sex: Male, 7 (58%); Female, 5 (42%) Cancer type: Unknown Cancer stage: Unknown Timing: After radiation therapy Treatment: Radiotherapy Intervention details: IG (Energy conservation): 1–2 h Face-to-face training and 3 telephone follow-ups ; CG (usual care): Standard care Research period: From 2 to 4 weeks after therapy Intervention frequency: One in-person training and weekly telephone follow-ups for 3 weeks Measurements: R-PFS Timing of rating: Preassessment within 2 weeks after therapy and postassessment after 3 telephone follow-ups Sample size: IG, 69; CG, 66 Age: IG, 55.7 ± 10.2 years; CG, 55.6 ± 13.7 years Sex: Male, 0 (0%); female, 135 (100%) Cancer type: Breast, 135 (100%) Cancer stage: IG: I, 13 (19%); Ⅱ, 28 (43%); and III, 25 (38%); CG: I, 13 (20%); II, 31 (44%); and III, 25 (36%) Timing: At the start of treatment Treatment: Chemotherapy Regimen: Unknown Intervention details: IG (Energy conservation training): Goal setting, activity prioritization, journaling, development of energy management skills education on work–life balance, discussion on resource use, and practice and homework to monitor their fatigue and activity CG: Standard care only and patients were placed on a waitlist Research period: 5 weeks Intervention frequency: 5 weekly 90-min group sessions Measurements: CFS Timing of rating: Preintervention, postintervention, and 8-week follow-up IG: CRF was reduced from pre- to postintervention and continued to decrease over the 8-week follow-up period (F = 69.8, P < 0.001) CG: The CFS levels showed no statistically significant increases over time Sample size: IG, 37; CG, 37 Age: IG, 50.1 ± 8.6 years; CG, 52.8 ± 10.1 years Sex: Male, 0 (0%); female, 74 (100%) Type of cancer: Breast, 74 (100%) Cancer stage: IG: I, 12 (32%); II, 17(46%); and III, 8 (21.6%); CG: I, 14 (38%); II, 17 (46%); and III, 6 (16%) Timing: During chemotherapy; Treatment: Chemotherapy; Regimen: AC regimen only Details of the intervention: IG (energy conservation program): Day 1: 30 min face-to-face health education in symptom management; days 2 and 7: 20 min telephone call assessment of self-care activity and monitoring CG: Energy conservation program in the study focused Research periods: 2 weeks Frequency of intervention: Days 1, 2, and 7 on AC therapy Measurements: NRS Timing of rating: Days 1 (baseline), 2, 7, and 14 Postintervention period: The median fatigue score in the IG improved significantly at the postintervention period compared with that in the CG. Cycle 1: The score was also significantly lower in the IG than that in the CG on day 14 (P < 0.01) Cycle 2: Fatigue scores on days 7 and 14 in the IG were significantly lower than those in the CG group (P = 0.02 and 0.01, respectively) Cycle 3: Only the fatigue score on day 7 in the IG was significantly lower than that in the CG (P = 0.04) Cycle 4: Fatigue scores on days 1, 2, 7, and 14 in the IG were significantly lower than those in the CG (P < 0.001, 0.04, 0.01, and 0.01, respectively) Sample size: IG, 25; CG, 25; Age: IG, 45.9 ± 7.6 years ; CG, 46.7 ± 1.7 years Sex: Male, 0 (0%); female, 50 (100%) Cancer type: Breast, 50 (100%) Cancer stage: IG: I, 7 (32%); II, 15 (68%). CG: I: 8 (38%); II, 13 (62%) Timing: During the third-to-fourth cycle of chemotherapy; Treatment: Chemotherapy Regimen: Unknown Intervention details: IG (energy conservation and concentrative movement therapy): Session-based energy management and self-assessment, fatigue assessment, general education on fatigue, and practice in energy conservation CG: Three 90-min weekly educational sessions on self-management during chemotherapy Research periods: 8 weeks ; Frequency of intervention: Weekly, three 90-min sessions Measurements: F-VAS, CFS, and R-PFS Timing of rating: Pretreatment, post-treatment, 1 week after treatment, and 4 weeks after treatment In all time points (after the last session, 1 week, and 4 weeks), F-VAS, CFS, and PFS scores significantly differed between the IG and CG (P < 0.001) All three CFS subscales (physical, affective, and cognitive) significantly decreased in the IG compared with those in the CG (P < 0.001) Sample size: IG, 10; CG, 9 Age: IG, 43.5 ± 9.1 years; CG, 48.8 ± 6.5 years Sex: Male, 0 (0%); female, 19 (100%) Cancer type: Breast, 19 (100%) ; Cancer stage: IG: I, 1 (10%); II, 8 (80%); and III, 1 (10%). CG: I: 1 (11%); II, 5 (56%); and III, 3 (33%) Timing: During chemotherapy Treatment: Chemotherapy Regimen: Unknown Intervention details: IG (ECAM): Prioritization, planning, stress management, and journaling. CG: Provision of general health information only Research period: 12 weeks ; Intervention frequency: Weekly Measurements: R-PFS; Timing of rating: Preintervention and 12 weeks Fatigue scores at 12 weeks were lower in the IG (X = 1.41, SD = 1.35) than in the CG (X = 3.32, SD = 1.42) The fatigue score from baseline to week 12 showed no significant difference within each group Independent t-test results revealed that the scores for fatigue intensity at week 12 significantly differed between the IG and CG (t[17] = − 3.006, P = 0.008) The effect size for fatigue change was determined using Hedges’ g (g = − 1.38; 95% CI [− 2.35, − 0.41]) Sample size: IG, 30; CG, 28 Age: IG, 40.2 ± 10.9 years; CG, 42.5 ± 12.3 years Sex: Male, 0 (0%) and Female, 58 (100%) Cancer type: Breast, 58 (100%) Cancer stage: IG: I, 3 (10%); II, 15 (50%); III, 11 (37%); and IV, 1 (3.3%). CG: I: 2 (7%); II, 10 (36%); III, 13 (46%); and IV, 3 (11%) Timing: During chemotherapy Treatment: Chemotherapy Regimen: Unknown Intervention details: IG (cognitive behavioral therapy with integrated activity pacing): SMART goal setting, relaxation techniques, and cognitive restructuring CG: Standard care only Research period: 7 weeks; Frequency of intervention: 3 face-to-face sessions and 4 phone sessions Measurements: BFI Timing of rating: Preintervention, postintervention (end of the intervention), and 3-month follow-up AC:, BFI, Brief Fatigue Inventory; CFS, Cancer Fatigue Scale; CG, control group; CRF, cancer-related fatigue; ECAM, energy conservation and activity management; F-VAS, Fatigue Visual Analog Scale; GFS, General Fatigue Scale; IG, intervention group; NRS, Numeric Rating Scale; POMS, Profile of Mood States; RCTs, randomized controlled trials; RM-ANCOVA, repeated measures analysis of covariance; R-PFS, Revised Piper Fatigue Scale; SCFS, Schwartz Cancer Fatigue Scale; SMART goals, Specific, Measurable, Achievable, Relevant, and Time-bound goals Table 4 Characteristics of eligible nonrandomized studies First author, year, and countryParticipantsCancer treatmentSummary of the interventionOutcome measureSummary of results Barsevick 2002, United States Sample size: IG, 38; CG, 182 Age: IG: radiotherapy, 59.4 ± 12.1 years; chemotherapy, 57.4 ± 13.4 years. CG: 55.1 ± 13.9 years Sex: Male, 84 (38%); female, 136 (62%) Cancer type: Breast, 89 (40%); gynecologic or lymphoma, 84 (38%); colon, 32 (15%); and lung, 15 (7%) Cancer stage: UnknownTiming: During chemotherapy (≥ 3 cycles of chemotherapy) or radiotherapy (≥ 6 weeks of radiotherapy) Treatment: Chemotherapy, 20 (53%); radiotherapy, 18 (47%) Regimen: UnknownIntervention details: IG (ECAM): Provision of information to patients, fatigue diary, activity prioritization list, and energy conservation planning. CG: Standard care only Research period: From treatment initiation to expected fatigue peak. Intervention frequency: Weekly telephone sessions (3 sessions)Measurements: POMS Timing of rating: Patients receiving chemotherapy: before the initiation (baseline), 48 h before second chemotherapy, 48 h after second chemotherapy. Patients receiving radiotherapy: before the initiation (baseline), last week of radiotherapy, 1 month after radiotherapy completionPatients receiving chemotherapy: The POMS fatigue scores tended to differ in the hypothesized direction between the IG and CG (F = 2.4; P = 0.12; df = 1.94). Patients receiving radiotherapy: CG exhibited higher POMS fatigue scores than the IG, indicating a trend toward between-group differences at both time points (F = 2.6; P = 0.11; df = 1.62) Getu 2022, Ethiopia Sample size: 10Age: 38.3 ± 7.2 yearsSex: Male, 0 (0%); female, 10 (100%)Cancer type: Breast, 10 (100%)Cancer stage: I, 1 (10%); II, 2 (20%); III, 5 (50%); and IV, 2 (20%)Timing: During chemotherapy (nonrelapsed in any disease stage)Treatment: ChemotherapyRegimen: UnknownIntervention details: IG (activity pacing [AP] used): SMART goal setting, relaxation techniques, and cognitive restructuringCG: Standard care onlyResearch periods: 6 weeks;Frequency of intervention: 6 sessions (3 face-to-face, 3 by phone)Measurements: BFI Timing of rating: Pretreatment (baseline) and at 6 weeks (end of the intervention)Fatigue severity (BFI) decreased in 77.77% of patients from a mean of 7.40 (± 2.21) at baseline to 3.50 (± 1.94) at 6 weeks of intervention (P < 0.05)Lu 2024, ChinaSample size: IG, 50; CG, 51Age: 21–70 yearsSex: Male, 0 (0%); Female, 101 (100%)Cancer type: Cervical, 101 (100%)Cancer stage: IG: I, 10 (20%); II, 23 (46%); III, 17 (34%). CG, I, 7 (14%); II, 23 (45%); and III, 21 (41%)Timing: Before treatment initiation Treatment: Concurrent chemoradiotherapy (radiotherapy: 5 times/week, chemotherapy: once in week 2) Regimen: UnknownIntervention details: IG: Knowledge provision about energy conservation, daily physical diary recording, and intervention with plan creation and adjustment. CG: Standard nursing care onlyResearch period: From treatment introduction to completion of 25 radiotherapy sessionsIntervention frequency: Weekly adjustments and intervention frequency stagedMeasurements: R-PFS;Timing of rating: Before treatment (T1); after 5, 10, 15, 20, and 25 radiotherapy sessions (T2, T3, T4, T5, and T6, respectively)According to the R-PFS (Chinese version) results, both the intervention and time main effects for the total score and for the behavioral, emotional, perceptual, and cognitive fatigue dimensions were statistically significant (P < 0.05)At T1 and T2, the total score and any fatigue dimension score did not significantly differ between the CG and IG (P > 0.05). Conversely, statistically significant differences were observed at T3, T4, T5, and T6 (P < 0.05)BFI, Brief Fatigue Inventory; CG, control group; ECAM, energy conservation and activity management; IG, intervention group; POMS, Profile of Mood States; RCTs, randomized controlled trials; R-PFS, Revised Piper Fatigue Scale; SMART goals, Specific, Measurable, Achievable, Relevant, and Time-bound goals Characteristics of eligible nonrandomized studies Sample size: 10 Age: 38.3 ± 7.2 years Sex: Male, 0 (0%); female, 10 (100%) Cancer type: Breast, 10 (100%) Cancer stage: I, 1 (10%); II, 2 (20%); III, 5 (50%); and IV, 2 (20%) Timing: During chemotherapy (nonrelapsed in any disease stage) Treatment: Chemotherapy Regimen: Unknown Intervention details: IG (activity pacing [AP] used): SMART goal setting, relaxation techniques, and cognitive restructuring CG: Standard care only Research periods: 6 weeks; Frequency of intervention: 6 sessions (3 face-to-face, 3 by phone) Sample size: IG, 50; CG, 51 Age: 21–70 years Sex: Male, 0 (0%); Female, 101 (100%) Cancer type: Cervical, 101 (100%) Cancer stage: IG: I, 10 (20%); II, 23 (46%); III, 17 (34%). CG, I, 7 (14%); II, 23 (45%); and III, 21 (41%) Timing: Before treatment initiation Treatment: Concurrent chemoradiotherapy (radiotherapy: 5 times/week, chemotherapy: once in week 2) Regimen: Unknown Intervention details: IG: Knowledge provision about energy conservation, daily physical diary recording, and intervention with plan creation and adjustment . CG: Standard nursing care only Research period: From treatment introduction to completion of 25 radiotherapy sessions Intervention frequency: Weekly adjustments and intervention frequency staged Measurements: R-PFS; Timing of rating: Before treatment (T1); after 5, 10, 15, 20, and 25 radiotherapy sessions (T2, T3, T4, T5, and T6, respectively) According to the R-PFS (Chinese version) results, both the intervention and time main effects for the total score and for the behavioral, emotional, perceptual, and cognitive fatigue dimensions were statistically significant (P < 0.05) At T1 and T2, the total score and any fatigue dimension score did not significantly differ between the CG and IG (P > 0.05). Conversely, statistically significant differences were observed at T3, T4, T5, and T6 (P < 0.05) BFI, Brief Fatigue Inventory; CG, control group; ECAM, energy conservation and activity management; IG, intervention group; POMS, Profile of Mood States; RCTs, randomized controlled trials; R-PFS, Revised Piper Fatigue Scale; SMART goals, Specific, Measurable, Achievable, Relevant, and Time-bound goals The included interventions could be broadly categorized as AP-related or energy conservation interventions. Five studies implemented interventions explicitly described as AP [12, 30] or activity management programs [10, 27, 29], whereas five studies evaluated energy conservation programs [11, 25, 26, 28, 31]. AP-related interventions primarily focused on self-regulation of activity patterns through goal setting, activity prioritization, and planning strategies. Conversely, energy conservation interventions emphasized techniques aimed at reducing energy expenditure, such as task simplification, planned rest periods, and fatigue management education. Despite these conceptual distinctions, substantial overlap in intervention components was observed across studies. Many programs incorporated similar strategies, including activity prioritization, planning daily activities, fatigue monitoring, and incorporating planned rest periods. Consistent with the conceptual framework described in the Background, these interventions share the common goal of regulating daily activity patterns and optimizing the use of limited energy resources to manage CRF. Therefore, for the purpose of this review, these approaches were collectively conceptualized as AP-related interventions, representing a broader behavioral energy management strategy. Considering these shared mechanisms, pooling the results across studies was considered conceptually appropriate. Risk of bias in studies Figure 2 shows a summary of the risk of bias assessment for the included studies. Four RCTs [11, 12, 27, 28] that employed computerized sequence generation or block randomization with allocation concealment had a low risk of bias in the randomization process (D1). In all studies, participants and care providers were aware of the assigned interventions due to the nature of the intervention program. Four trials [11, 12, 25, 27], which did not clearly deviate from the intended intervention, had low risk of bias in deviations from the intended intervention (D2). Two trials [26, 28] with a significant number of dropouts may have influenced the results, and one trial [12] that did not report baseline data was at high risk of bias for missing outcome data (D3). Because all trials used patient-reported outcomes and patients were aware of the intervention, all were judged to have some concerns regarding bias in outcome measurement (D4) [10–12, 25–28]. One study [26], which assessed several multiple time points without prespecifying them, and two studies [10, 28], which did not report prespecified analysis plans, were judged to have a high risk of bias for selection of the reported result (D5). In summary, four trials had a high overall risk of bias and three trials had some concerns of bias. Risk of bias assessment of the included randomized controlled trials (RCTs) Figure 3 shows a summary of the risk of bias assessment for the nonrandomized studies. Two trials [29, 30] that did not control for key confounding factors were judged to have a serious risk of bias because of confounding, and one trial [31] with suspected substantial unmeasured confounding was judged to have a critical risk of bias because of confounding (D1). All three trials [29–31], in which participants were assigned to groups before receiving the intervention with clear criteria, were judged to have a low risk of bias in both classifications of interventions (D2) and selection of participants into the study (D3). All three trials [29–31] had a moderate risk of bias because of deviations from the intended interventions, as they did not account for the effect of small deviations on the results (D4). One trial [31] with a substantial number of dropouts may have affected the results because of a critical risk of bias due to missing data (D5). Because all trials [29–31] used patient-reported outcomes and participants were aware of the intervention received, all were judged to have a serious risk of bias in outcome measurement (D6). Because no prespecified analysis plans were available for the three trials with multiple outcome measurements [29–31], they were judged to have a serious risk of bias in the selection of reported results (D7). In summary, one trial had a critical overall risk of bias and two trials had a serious overall risk of bias. Risk of bias assessment of the included nonrandomized controlled trials (non-RCTs) Effectiveness of AP-related interventions on CRF We performed a meta-analysis of the primary outcomes from five RCTs, as detailed data were unavailable for two RCTs (Fig. 4). Compared with the controls, the effect size of the intervention was small (SMD = − 0.36; 95% CI: − 0.73 to − 0.00; P = 0.05; five studies; 617 participants). Additionally, the heterogeneity index indicated substantial heterogeneity (I2 = 66%). The funnel plot did not suggest publication bias (Supplemental Fig. 1). However, because fewer than 10 studies were included in the meta-analysis, formal assessment of publication bias was not performed. Forest plots of the effect of activity pacing-related interventions on cancer-related fatigue Because of substantial heterogeneity among studies involving participants undergoing chemotherapy [11, 27, 28] and those involving participants undergoing or having completed radiotherapy [10, 25], subgroup analyses were conducted for each group. Compared with controls, the effect size of the intervention was moderate (SMD = − 0.58; 95% CI: − 0.87 to − 0.29; P < 0.01; three studies; 197 participants), with low heterogeneity (I2 = 0%) in the subgroup of patients undergoing chemotherapy (Fig. 4), whereas the effect size of the intervention compared was not significantly different (SMD = − 0.01; 95% CI: − 0.20 to 0.18; P = 0.94; two studies; 420 participants) and showed low heterogeneity (I2 = 0%) in the subgroup of patients undergoing or having completed radiotherapy. Leave-one-out sensitivity analyses revealed some variability in the pooled estimates, reflecting the small number of included studies; however, the overall direction of the effect remained consistent. Certainty of the evidence Figure 5 shows a summary of the certainty of evidence from the RCTs. Among the five RCTs, the certainty of the evidence was extremely low. The certainty of evidence was low in studies involving participants undergoing chemotherapy and those involving participants undergoing radiotherapy. Summary of randomized controlled trial findings on the certainty of evidence Additional outcomes reported in the included studies Several studies assessed outcomes beyond CRF, including functional performance, health-promoting behaviors, pain, sleep, depression, physical activity, and quality of life [10–12, 26, 30]. One study reported no significant improvement in overall functional performance, although the intervention group experienced less disruption of usual activities [10]. Improvements in health-promoting lifestyle behaviors were observed in another study [11]. Reductions in pain and depressive symptoms along with improvements in sleep and physical activity were also recorded [26]. Studies evaluating psychological outcomes and quality of life reported reductions in depressive symptoms and improvements in global health status and functioning alongside decreases in several symptom scales [12, 30]. Several studies did not report outcomes beyond fatigue (Supplemental Table 2) [25, 27–29, 31]. Components and BCTs included in AP-related interventions for CRF The primary BCTs used in AP-related interventions were “goal setting (behavior),” “social support,” and “graded tasks.” Additionally, frequently used BCTs included “information about behavior change strategies,” “self-monitoring,” and “managing negative emotions” (Fig. 6). Frequency of use of behavior change techniques in activity pacing-related interventions Discussion This systematic review and meta-analysis identified three principal findings. First, AP-related interventions tended to ameliorate CRF, although the pooled effect was not statistically significant, and the certainty of evidence was insufficient because of the high risk of bias and substantial heterogeneity. Second, AP-related interventions appeared more beneficial for patients undergoing chemotherapy; however, the certainty of evidence was low because of a high risk of bias, small sample size, exclusive focus on breast cancer, and limited number of available studies. Third, the key BCTs for AP-related interventions included “goal setting,” “social support,” and “graded tasks.” Although the meta-analysis indicated that AP-related interventions may improve treatment-related CRF, the pooled effect was not statistically significant and showed substantial heterogeneity. Notably, substantial heterogeneity was observed among studies involving patients undergoing chemotherapy [11, 27, 28] and those involving patients undergoing or having completed radiotherapy [10, 25]. This heterogeneity may be attributed to the radiotherapy studies being conducted nearly 20 years ago. The timing of AP-related interventions differed (during versus after chemotherapy and radiotherapy). Additionally, some studies did not specify cancer stage, potentially resulting in the inclusion of patients at different stages of disease progression [10, 25]. Furthermore, the relatively large sample size in the study by Barsevick [10] may have disproportionately influenced the results and contributed to the observed heterogeneity. In the ad hoc subgroup analysis of patients undergoing chemotherapy, AP-related interventions were associated with a significant improvement in moderate fatigue, specifically CRF secondary to chemotherapy. However, certainty of the evidence was extremely low because of the relatively small sample size, limited disease focus (studies targeting only breast cancer), limited number of studies, and wide CIs. Therefore, interpretation requires caution, and further well-designed studies are needed. In contrast, AP-related interventions did not improve CRF in the subgroup of patients who were undergoing or having completed radiotherapy. In addition to the high heterogeneity of the studies, their focus on energy conservation rather than maintaining physical activity may have affected their results [10, 25]. Sensitivity analyses indicated instability of the pooled estimates, likely attributable to the limited number of studies and small sample sizes. Therefore, the robustness of the observed effects should be interpreted cautiously. AP-related interventions ameliorate cognitive or psychosocial fatigue by reinforcing energy-conserving and valued activities and may also address physical fatigue by modulating physical activity through graded activity quotas [9, 11, 28]. However, it remains unclear whether these interventions result in sustained improvements in physical activity following the intervention and reduce fatigue. Notably, none of the studies in this review reported changes in physical activity levels, warranting further research. The present findings indicate that interventions incorporating AP or energy conservation strategies may contribute to reductions in CRF. Although the included studies differed in terminology and delivery formats, many shared common behavioral components, including activity prioritization, goal setting, fatigue self-monitoring, and strategic distribution of activity and rest. These elements align with behavioral fatigue management models designed to prevent boom–bust cycles of overexertion followed by symptom exacerbation. By facilitating improved regulation of daily activities and more efficient energy allocation, such strategies may stabilize activity patterns and reduce symptom exacerbation. Thus, AP and energy conservation interventions may operate through similar self-regulatory mechanisms associated with activity and energy expenditure. However, historical differences in terminology, structure, and delivery may have contributed to heterogeneity across studies; therefore, integrating these interventions within a single meta-analysis may introduce additional variability. Future studies should provide more standardized descriptions and clearer reporting of intervention components to distinguish the specific effects of different behavioral energy management strategies. As key BCTs, AP-related interventions in each study [10–12, 25–31] included “goal setting,” “social support,” and “graded tasks,” which may enhance intervention adherence and effectiveness. By incorporating and practicing AP-related interventions in daily life, patients can self-manage fatigue, thereby reducing its severity. Thus, psychoeducation should effectively enhance self-efficacy, promote behavioral change, and enable the practice of AP-related interventions. AP-related interventions should also include goal setting, social support, and graded tasks as they improve motivation [32]. In addition to effects on CRF, several studies reported potential improvements in additional patient-reported outcomes, such as depressive symptoms, physical symptoms, health-promoting behaviors, and health-related quality of life [10–12, 26, 30]. Such benefits may reflect more adaptive regulation of daily activities and energy expenditure. However, only a limited number of trials assessed these outcomes, and measurement variability across studies limited interpretability. Further well-designed studies are needed to clarify the broader impact of behavioral energy management interventions among cancer populations. The participants in this study were limited to those receiving chemotherapy. Although CRF is also prevalent among patients with terminal disease, no study has evaluated the effects of AP-related interventions in this population. Therefore, our results should be interpreted strictly within the context of patients undergoing chemotherapy. Future studies should investigate whether AP-related interventions are feasible and effective in more vulnerable populations, including those with advanced or terminal disease. The strength of this study is its comprehensive review of the effects of AP-related interventions on CRF. However, this study had several limitations. First, the included interventions varied in terminology and structure, and AP and energy conservation were synthesized within a broader category of behavioral energy management strategies. Second, a high risk of bias was observed in each study. All five RCTs that assessed the efficacy of AP-related interventions on CRF had a high risk of bias. Although several RCTs were well-designed, the use of self-reported outcomes and nature of AP-related interventions as a psychoeducational intervention prevented the blinding of both investigators and participants, resulting in a high risk of bias. Third, because different instruments were used to measure CRF across studies, the pooled effect was calculated using SMDs, which may limit the clinical interpretability of the magnitude of the observed effects. Fourth, we could not contact some of the original authors; this review may be updated in the future if additional data become available. Fifth, formal independent parallel coding and inter-rater reliability assessment were not conducted. Although final BCT assignments were confirmed by consensus among all four authors for each study, future studies should consider independent double coding by trained coders, quantification of inter-rater agreement, and a prespecified adjudication procedure to further strengthen methodological rigor and reproducibility. Sixth, publication bias could not be formally assessed because of the small number of included studies. Therefore, the possibility that the observed effects were influenced by unpublished or selectively reported studies cannot be excluded. Although the pooled effect for CRF was small (SMD = − 0.36), even modest improvements may be clinically meaningful, as small reductions in fatigue can improve daily functioning and quality of life. Considering that these interventions are generally low risk, even small benefits may be clinically meaningful. However, the clinical relevance remains uncertain considering the lack of statistical significance, the confidence interval including the null value, and the absence of well-established minimal clinically important differences for CRF across outcome measures. Conversely, the moderate and significant effect observed in the chemotherapy subgroup (SMD = − 0.58) suggests that AP-related interventions may be particularly beneficial for treatment-related CRF. Despite these encouraging findings, the overall certainty remains extremely low. Future research should standardize intervention components and outcome measures to strengthen the evidence base and clarify the mechanisms underlying these interventions. Future studies should address the following gaps: (1) conduct large-scale, rigorously designed RCTs to confirm the effectiveness of AP-related interventions; (2) systematically assess their applicability to populations beyond those undergoing chemotherapy, such as patients receiving radiotherapy or those with advanced disease; and (3) incorporate objective measures of physical activity and functional outcomes in addition to self-reported fatigue scales to better elucidate the mechanisms underlying AP-related interventions. Effective AP-related intervention programs include key BCTs such as “goal setting,” “graded tasks,” and “self-monitoring.” Digital technology can enhance these BCTs by integrating data from wearable sensors and patient-reported fatigue, enabling real-time adjustments of activity–rest balance. These digitized AP-related intervention programs may facilitate future large-scale studies and their implementation, which should be guided by a digital ontology of BCTs [33]. Conclusion AP-related interventions may provide some benefit in improving CRF, particularly during chemotherapy. However, the certainty of the current evidence remains extremely low because of a high risk of bias, limited disease focus (primarily breast cancer), small sample sizes, substantial heterogeneity, and a limited number of studies. Thus, definitive conclusions cannot yet be drawn. Although AP-related interventions may be considered a potentially useful approach, high-quality RCTs are warranted in the future. Clinically, AP-related interventions are sometimes implemented in vulnerable populations, such as patients with terminal disease; thus, future research should also focus on these populations. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 186 kb) Acknowledgements The authors would like to thank Enago (www.enago.jp) for the English language review. Author contribution YS conceived and managed the entire research project. YS, KT, AH, and MH wrote the protocol for this study, analyzed, and interpreted the data. YS and MH screened the literature detected in the databases. YS, KT, and MH evaluated the risk of bias of the eligible studies. YS drafted the manuscript. YS, KT, AH, and MH critically revised the manuscript for important intellectual content. YS obtained funding. All authors have read and approved the final manuscript. Funding This work was supported by JSPS KAKENHI Grant JP24K22219 and by Secom challenge 2023 Research grant from Secom Science and Technology Foundation (AH). The funders played no role in the study design, data collection and analysis, decision to publish, or manuscript preparation. Data availability The data supporting the findings of this study are derived from previously published studies and are available in the cited articles. The extracted data used for the analyses are available from the corresponding author upon reasonable request. Declarations Ethical approval Ethical approval was not required. The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary file1 (PDF 186 kb) Data Availability Statement The data supporting the findings of this study are derived from previously published studies and are available in the cited articles. The extracted data used for the analyses are available from the corresponding author upon reasonable request.