This meta-analysis was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 statement [38] and has been duly registered with the International Prospective Register of Systematic Reviews (PROSPERO; registration number CRD42024582905). As the data utilized in this research were exclusively sourced from previously published studies, ethical approval and informed consent were not required.

After consulting a professor of statistics, a comprehensive literature search was carried out, covering the period from the inception of the databases to July 31, 2024. The search was conducted across PubMed, Embase, Web of Science, CENTRAL, and MEDLINE, with access to the full text of the articles. The development of the search strategy was guided by the PICOS (Participants, Interventions, Comparisons, Outcomes, and Study Design) framework, along with the guidelines provided by the Cochrane Collaboration to ensure the integrity of the analysis. This strategy incorporated the use of MeSH (Medical Subject Headings) terms, textual keyword searches, and Boolean logic operations, supplemented by keywords from titles or abstracts, including terms such as neoplasms, carcinomas, tumors, cancer, caregivers, PA trackers, wearable, telemedicine, and telerehabilitation. All search strategies used are presented in Multimedia Appendix 1. The search was limited to studies involving humans and randomized controlled trials and was conducted in English. Moreover, we conducted a rigorous manual review of the bibliographies of the retrieved articles to identify and obtain supplementary relevant scholarly works, thereby enhancing the depth of our analytical inquiry.

This meta-analysis considered studies for inclusion based on the following criteria: (1) participants were survivors of any type of cancer, regardless of sex or cancer stage; (2) patients in the intervention group received WEDS-supported PA programs, which included reminders to change behavior, consultations with HCPs, or social support from other patients; (3) patients in the control group received usual care or were placed on a waitlist; (4) the outcomes included at least one of the following indices—objectively measured MVPA, subjectively reported PA, sedentary behavior, QoL, or BMI—without restrictions on the measures used; (5) the publications were written in English; and (6) the studies were designed as randomized controlled trials. Studies were excluded if they were only registered but not yet conducted or if relevant data were incomplete.

The reference management software EndNote X9 (Clarivate Plc) was used to import and screen the titles and abstracts of the studies. Duplications were first removed automatically by EndNote and then meticulously screened by researchers. To ensure alignment with the inclusion criteria, 2 independent authors (ZW and YL) concurrently conducted a thorough screening of the titles and abstracts. Subsequently, they carefully evaluated the full texts of the papers based on the predetermined eligibility criteria. Any discrepancies in the screening process were resolved through discussion or by consulting a third author (QW). Data extraction from the included studies was performed independently by 2 authors (ZW and YL), who meticulously recorded the information using a predefined data extraction template. This template encompassed a range of details, including the first author’s name, year of publication, country where the study was conducted, participants’ ages (mean and SD), sample size, type of cancer diagnosed, types of WEDs and associated tools used, intervention content, duration of the interventions, outcome measures employed, and timing of assessments.

The methodological quality and risk of bias in the included studies were meticulously assessed by 2 independent reviewers, using the Risk of Bias Tool 2, version 5.1.0. A total of 7 domains were evaluated: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other potential sources of bias. Each domain was graded as “low risk” of bias, “high risk” of bias, and “unclear risk” of bias. The official Cochrane Excel tool was used to automatically compute the overall risk. Disagreements were resolved through discussion or, when necessary, by consulting a third author.

Utilizing R Studio (R Foundation), we conducted heterogeneity evaluations and performed the meta-analysis. To quantify the intervention effects, we computed the standard mean difference (SMD) along with its corresponding 95% CI, and presented the results using forest plots. To obtain more robust results, all data were pooled and analyzed using a random-effects model, while a fixed-effects model was applied when the number of included studies was small (no more than 5) [39]. In cases where a multiarm trial was included, the shared group was divided into subgroups of approximately equal size, 1 for each experimental group [40,41]. In addition, we assessed statistical heterogeneity across all included studies using the I² statistic and P value. When there were 10 or more studies, the Egger test was conducted to assess small-study effects, with a P value below .05 indicating the possible presence of such effects [42]. To evaluate the robustness and reliability of the pooled results, a sensitivity analysis was performed, using the one-study-out method. Statistical significance for the overall effect was established when the 2-tailed P value was less than .05.

The initial search across 5 electronic databases identified 4555 articles. After 1194 duplicates were removed both automatically and manually, 3361 articles were excluded based on their titles and abstracts. Following this initial screening, the full texts of the remaining 153 articles were retrieved, resulting in a final total of 46 studies included in the meta-analysis. The procedures for search and selection are delineated in Figure 1.

Researchers in 12 studies reported feasibility, which was assessed by retention rate (n=8), wearing time (n=2), whether steps per day improved or not (n=1), and adherence to interventions (n=1) [39-41,44,45,48,49,59,62,63,67,69,77,78].

Utilizing the revised Cochrane risk-of-bias tool, the 24 studies that utilized intention-to-treat analysis within the inclusion criteria were classified as follows: 6 (25%) studies [44,51,55,77,78,84] were deemed to have a low risk of bias, while 18 (75%) studies [43,46,48,54,57,59,61,62,67,70,71,74,76,80,82,85,86,88] were identified as having some concerns regarding bias. Furthermore, among the studies that used per-protocol analysis (N=22), 6 (27%) studies [47,49,64,75,79,81] were assessed as having a low risk of bias, whereas 15 (68%) studies [50,52,53,56,58,60,63,65,66,68,69,72,73,83,87] were classified as having some concerns, and 1 (5%) study [45] was classified as having a high risk of bias. Concerns regarding risk of bias emerged due to the randomization process (33/47 studies) [43,46,48,50,52-54,56-63,65-74,76,80,82,83,85-88] and the measurement of the outcome (1 of 47 studies) [52]. A high risk of bias was associated with deviation from the intended interventions (1 of 47 studies) [45]. The assessments of risk of bias are comprehensively presented in Figure 2. In addition, the results of the Egger test revealed no evidence of small study effects (objectively measured MVPA: P=.26; subjectively reported PA: P=.09; steps per day: P=.12; sedentary behavior: P=.15; BMI: P=.13; QoL: P=.24; Multimedia Appendix 3).

The summary of all outcomes included in this meta-analysis is detailed in Multimedia Appendix 3.

A total of 46 studies met the eligibility criteria for this meta-analysis. Compared with usual care or waitlists, WEDS-supported PA programs significantly improved cancer survivors’ objectively measured MVPA, subjectively measured PA, steps per day, and QoL, but showed no significant effect on reducing sedentary behavior or improving BMI. The SMD of total effects ranged from −0.07 to 0.66, which is consistent with previous studies [23,29], thereby confirming the effectiveness of WEDS-supported PA programs.

Overall, the evidence quality and methodology were rated as moderate; 24 studies used the intention-to-treat analysis, while 22 utilized per-protocol analysis. Using the revised Cochrane Risk-of-Bias tool, it was determined that 12 of 46 (26%) studies were assessed as having a low risk of bias, 33 of 46 (72%) studies raised some concerns, and only 1 of 46 (2%) studies was deemed to have a high risk of bias. Specifically, of these 46 studies, 33 (72%) were flagged for potential bias related to the randomization process, and 1 (2%) raised concerns regarding bias in the measurement of outcomes. In detail, all studies were assessed as having either a low or unclear risk associated with the randomization process, primarily because some investigators failed to provide comprehensive details on the randomization techniques or adequately describe the allocation concealment methods. Consequently, the overall methodological quality was deemed moderate. These findings underscore the need for additional randomized controlled trials in the future, with a focus on more transparent reporting to enhance the robustness of research findings.

The results of this meta-analysis demonstrated that WEDS-supported PA programs significantly improved objectively measured MVPA, subjectively reported PA, steps per day, and QoL in cancer survivors, but had no significant effect on sedentary behavior or BMI. The findings related to objectively measured MVPA, subjectively reported PA, and steps per day are consistent with those of previous studies [23].

For cancer survivors, these interventions serve as tracking devices (continuously collecting current activity), feedback tools (providing immediate information on activity levels), and environmental cues (reminders to be active). Through these continuous influences, patients’ levels of PA increase [91]. In addition, the partnering tools used in these interventions enable cancer survivors to record, report, and contact HCPs anytime and anywhere they need, ensuring timely revision of exercise prescriptions and providing knowledge related to their disease and symptoms [78,80]. Moreover, the similar results between subjectively reported PA and objectively measured MVPA highlight the feasibility of WEDS-supported PA programs in improving PA among cancer survivors. On the contrary, we found that sedentary behavior did not improve significantly. This may be because patients may choose to ignore activity alarms while remaining sedentary [46]. More cognitive behavioral therapy is needed to enhance patients’ awareness and motivation to reduce sedentary behavior [92]. Regarding BMI, the lack of a significant difference might be due to patients gaining muscle mass, which can offset weight loss, resulting in no apparent change in BMI despite positive effects on fitness [93]. Furthermore, diet management has been shown to play a more significant role in weight loss programs than PA alone in many studies [94,95].

The results of the subgroup analysis on the usage of multipartnering tools revealed that objectively measured MVPA significantly improved regardless of whether multipartnering tools were used. This outcome may be because, at the beginning of the interventions, researchers set PA goals for participants and adjusted those goals based on their performance through the partnering tools. Whether or not multipartnering tools were used, the partnering tools could still serve as reminders for patients to complete more MVPA. Thus, the use of multipartnering tools may not have influenced objectively measured MVPA. By contrast, for subjectively reported PA, a significant difference was observed in the nonusage of multipartnering tools. When patients are assisted by multipartnering tools to remind them to improve PA, they may lack initiative, and their self-efficacy regarding PA may decrease, along with their perception of subjectively reported PA [96]. For steps per day, the multipartnering tools groups showed significant differences. This may be because various partnering tools play a more comprehensive role, such as providing real-time conversations through telephone calls and offering relevant knowledge via apps or websites. Through this type of multimedia stimulation, participants can receive more comprehensive reminders and encouragement to remain physically active.

Moreover, we observed significant improvements in objectively measured MVPA, subjectively reported PA, steps per day, and QoL among participants who received long-term (no less than 12 weeks) interventions, which is consistent with previous similar findings [36]. Research has shown that longer-term interventions are conducive to forming healthier lifestyles and developing lasting habits. Therefore, the patients can derive enjoyment from PA and are more willing to complete additional PA programs [97,98]. Similar to the overall effects, the results showed no significant improvements in sedentary behavior or BMI in the subgroups categorized by duration.

When studies were grouped by whether the intervention was designed for a specific cancer type, the results showed that both steps per day and QoL significantly improved in patients who received interventions tailored to their specific cancer type. This may be because tailored programs could better meet patients’ specific needs, such as providing information related to their cancer type and offering PA programs suited to their condition. By contrast, regardless of whether the interventions were tailored or not, the effectiveness of WEDS-supported PA programs in improving objectively measured MVPA, subjectively reported PA, sedentary behavior, and BMI did not change.

In brief, WEDS is promising for improving MVPA, subjectively reported PA, steps per day, and QoL. Long-term interventions (≥12 weeks) are effective in improving PA-related outcomes, except for sedentary behavior, and the use of multipartnering tools should depend on the patients’ preferences and habits. Attention should also be given to the proper use of multipartnering tools, the optimal duration, and whether the intervention is tailored to specific cancer types when developing new WEDS-supported interventions. Further studies are needed to explore the most effective intervention characteristics for improving patients’ sedentary behavior and BMI. This approach will facilitate the development of more effective WEDS-supported interventions.

This meta-analysis demonstrated that, compared with usual care or waitlists, WEDS-supported PA programs have a significant effect on the QoL of cancer survivors, which is consistent with the findings of previous studies [23]. The QoL assessed in the included studies was health-related QoL, which encompasses not only basic physical functioning but also patient participation in activities such as work and entertainment [99]. WEDS-supported PA programs significantly improved cancer survivors’ inactive lifestyles, enhanced their self-efficacy and feelings of self-worth, increased their satisfaction with life, and indirectly influenced their QoL [100]. Moreover, appropriate social relationships, cancer and self-care education, and psychological support provided through partnering tools could further help improve cancer survivors’ QoL [101,102].

In the subgroup analysis, regarding the use of multipartnering tools, QoL was significantly improved in both the usage and nonusage groups. This may be because interventions in both groups provided reminders to patients, which significantly enhanced participants’ PA levels and indirectly reduced their symptom burden, thereby improving QoL. When grouped by intervention duration, QoL was significantly improved in the long-term subgroups. Longer intervention durations enable patients to develop sustained habits of positive PA. Additionally, patients may have more opportunities to access diverse forms of support over the long term, which provides greater encouragement for engaging in PA and fosters the adoption of self-management strategies, thereby improving their QoL [103]. Furthermore, we observed a significant improvement in QoL among patients who received interventions designed for a specific cancer type. Researchers could tailor specific programs, such as PA regimens and psychological support from HCPs, according to the characteristics of each patient’s cancer.

In essence, WEDS-supported PA programs enable cancer survivors to engage positively with WEDs and partnering tools, with the potential to reduce negative affective states and consequently enhance their QoL. Researchers should carefully consider the duration of intervention when designing WEDS-supported strategies. Additionally, further investigation is warranted to evaluate the effectiveness of partnering tools in addressing the specific needs of cancer survivors.

This study has several limitations. First, heterogeneity existed due to variations in the format of partnering tools, durations of intervention, and types of cancer. Second, the reporting of study results may have been influenced by commercial interests associated with PA improvements, posing a potential risk of publication bias. Additionally, a significant proportion of the research was conducted in Western countries, and responses to WEDS-supported PA programs may vary among participants from different regions [104]. Finally, despite conducting an exhaustive literature search, publication bias could not be completely eliminated. Therefore, the outcomes of this meta-analysis should be interpreted with caution, and more high-quality randomized controlled trials are needed in the future.

In this study, we quantitatively integrated existing findings and found that WEDS-supported PA programs were effective in improving PA levels (both objectively and subjectively), daily steps, and QoL. The mechanisms through which WEDS-supported PA programs bring clinical benefits may include providing persistent reminders to encourage PA, offering convenient access to consultations with HCPs, collecting health-related data, recording electronic health records, and facilitating social groups for patients to communicate with others facing similar conditions [47,50]. Thus, HCPs can use WEDS as a supplementary tool to monitor patients’ physiological data, manage care, adjust exercise prescriptions, and provide timely feedback and disease-related information.

With increasing research focusing on WEDs and other forms of eHealth as interventions to promote PA among cancer survivors, WEDS has the potential to become a valuable tool for HCPs and a novel reminder and management resource for cancer survivors. It can automatically sync data, thereby reducing the self-monitoring burden associated with traditional web-based interventions [105]. Additionally, previous studies often failed to adequately consider the role of partnering tools, resulting in their underutilization and a missed opportunity to maximize the benefits for patients’ PA engagement. Furthermore, we observed that certain aspects of the intervention, such as the use of multipartnering tools, the duration of the intervention, and whether the intervention was tailored for specific cancer types, influenced its overall efficacy. This underscores the need for further standardization and more rigorous quantitative studies to refine the WEDS-supported intervention framework and to fully explore the potential benefits of WEDS-supported PA programs. Moreover, efforts should be made to enable data intercommunication between different commercial WEDs, thereby improving the feasibility and accessibility of these interventions.

WEDS-supported PA programs offer a convenient and affordable method for assisting cancer survivors by serving as reminders and records of their PA. This meta-analysis of randomized controlled trials revealed that WEDS-supported PA programs significantly improved cancer survivors’ level of PA (both objectively and subjectively), steps per day, and QoL, but had no significant effect on reducing sedentary behavior or BMI. These results varied based on the use of multipartnering tools, intervention duration, and patients’ cancer type. Further standardization and promotion of WEDS-supported PA programs are warranted in the future.