Kinect-Based Mixed Reality Exercise Program Improves Physical Function and Quality of Life in Breast Cancer Survivors: A Randomized Clinical Trial

Article information

J Korean Cancer Assoc. 2024;.crt.2024.758

Publication date (electronic) : 2024 October 30

doi : https://doi.org/10.4143/crt.2024.758

Byunggul Lim^1,2,3, Xinxing Li¹, Yunho Sung¹, Parivash Jamrasi¹, SoYoung Ahn¹, Hyejung Shin¹, Wook Song^,1,2,4,5

¹Health and Exercise Science Laboratory, Department of Physical Education, Seoul National University, Seoul, Korea

²Institute on Aging, Seoul National University, Seoul, Korea

³Director of Research, Dr. Exsol, Inc. Research Institute, Seoul, Korea

⁴Institute of Sport Science, Seoul National University, Seoul, Korea

⁵Chief Executive Officer, Dr. Exsol, Inc. Research Institute, Seoul, Korea

Correspondence: Wook Song, Health and Exercise Science Laboratory, Department of Physical Education, Seoul National University, 410, 71-1, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Korea Tel: 82-2-880-7791 E-mail: songw3@snu.ac.kr

Received 2024 August 8; Accepted 2024 October 28.

Abstract

Introduction

In 2023, with 2.3 million new cases, constituting 12.5% of all cancer diagnoses, breast cancer in women has surpassed lung cancer as the foremost contributor to the global cancer incidence [1] and, with 685,000 fatalities, it is the fifth most prevalent cause of cancer-related deaths [2]. Exercise is an effective non-pharmacological intervention for breast cancer survivors (BCS) to alleviate treatment-related adverse effects, enhance quality of life (QOL), improve cardiorespiratory fitness, and increase muscular strength [3]. Combined aerobic and resistance exercises can improve the QOL and attenuate adipose tissue inflammation in obese BCS [4]. Breast cancer-related lymphedema (BCRL) is among the most common adverse effects related to cancer treatment [5]. Also, 16 weeks of resistance and high-intensity interval training (HIIT) was effective in preventing increases in cancer-related fatigue and in reducing symptom burden for patients during chemotherapy for breast cancer [6]. Systematic reviews have indicated that stretching, yoga, qigong, and Pilates are safe, effective approaches for BCRL risk or symptom management [7].

Nonetheless, traditional exercise programs have limitations such as the lack of adherence or monitoring of exercise intensity and progression, as well as the limited access to exercise facilities. Barriers to exercising in cancer survivors include time and cost [8], lack of enjoyment or equipment [9], psychological distress, and lack of motivation or confidence [10]. Therefore, BCS frequently seek exercise alternatives in a safer, familiar environment [11].

Several exercise intervention studies have aimed to address the limitations of traditional exercises in BCS. Exergaming confers functional upper limb improvement and enhanced daily task performance [12]. Despite the abundance of research on conventional exercise methods, there is a noticeable gap in studies focusing on the use of advanced technologies, such as virtual reality (VR), mobile applications, and other digital platforms, for exercise interventions in breast cancer patients. Although not many studies have utilized VR, some have shown that VR training, which includes body-weight exercises and active motions of the upper limb such as all movements of the shoulder joint, wrist and fingers, was found to be superior to resistance exercise training in managing BCRL. It showed significant improvements in shoulder range of motion (ROM), body pain, general health, vitality, and Disabilities of the Arm, Shoulder, and Hand (DASH) scores [13,14]. With a primary focus on gaming, exergaming and VR-based exercises are not specifically developed for BCS, and there are low adherence rates and a higher likelihood of incorrectly performing exercises when following instructional exercise videos [15]. Accurate assessment of exercise intensity and improvements are hindered by lack of user data and records, even with specific device use. Therefore, Kinect-based mixed reality devices (KMR) comprise three components: the Kinect device, for real-time movement tracking of 25 joints; the main screen, which displays the user’s image; and the floor environment, for interaction between the user and augmented reality (Fig. 1).

Fig. 1.

Kinect-based mixed reality device.

KMR exercise programs incorporate gaming elements and provide real-time feedback, with user data collection features that enable users to properly exercise while monitoring individual exercise intensity and progress; furthermore, preserved user data facilitates outcome assessment [16].

The effectiveness of exercise intervention using the KMR system in improving physical function, alleviating fatigue, addressing upper extremity dysfunction, and enhancing QOL in BCS has not been elucidated. A deeper understanding of the health-related impacts of KMR exercise programs will enable targeted exercise intervention recommendations for BCS.

This study aimed to elucidate the effects of the KMR exercise program on physical functioning, lymphedema symptom severity, and QOL in BCS. We posited that, compared to the home stretching group (HS), an 8-week KMR exercise program would greatly improve physical function (shoulder ROM, hand grip strength [HGS], and aerobic capacity), QOL, upper extremity disability, and physical activity (PA).

Materials and Methods

1. Research design

Participants were recruited through local cancer-related associations, social media advertisements, and referrals from the Korean Cancer Society (Seoul, Korea) and randomly assigned to two groups (Fig. 2). The inclusion criteria were as follows: (1) a confirmed diagnosis of stage I-III breast cancer; (2) completion of radiotherapy and/or chemotherapy at ≥ 6 months before study enrollment; (3) no metastasis or other cancers; (4) engagement in physical activities for < 60 minutes per week, including resistance exercise, within the preceding 6 months; (5) absence of musculoskeletal, neurological, metabolic, cardiovascular, and respiratory diseases; and (6) no contraindicated medications or comorbidities hindering participation in exercise programs. Individuals with severe cardiac disease, uncontrolled hypertension, lymphedema (intact limb difference > 2 cm), neurological or rheumatic disease, communication difficulties, recurrent infection, open wounds, or abdominal incisions were excluded, as were those who underwent continuous chemotherapy or radiotherapy.

Fig. 2.

Experimental design. QOL, quality of life; ROM, range of motion.

The sample size of 66 was calculated using G*Power 3.1, with an effect size of 0.2, power 0.95, and significance level (α) 0.05. To account for a projected dropout rate of 20%, eighty participants were enrolled, of which three dropped out because of reasons related to distance. The final sample for analysis included 77 participants, who were allocated to the Kinect-based group (KE) and HS (38 and 39, respectively) using a computer-generated randomization sequence to maintain concealment from the study investigators. Prior to the interventions, all participants underwent pre-test assessments of HGS, Chester step test (CST), shoulder ROM, questionnaires (Korean version of DASH [K-DASH], Korean version of the International Physical Activity Questionnaire [K-IPAQ], and Functional Assessment of Cancer Therapy-Breast [FACT-B]), and body-composition analysis.

2. Interventions

The KMR system integrates mixed reality to provide real-time performance feedback and track adherence, allowing participants to view their progress on-screen and receive immediate feedback on movement quality. This feedback ensures correct form and motivates participants to stay engaged throughout the program. The system automatically adjusts exercise intensity by increasing repetition targets based on performance, ensuring personalized progression. By using HIIT, the program maximizes cardiovascular and muscular endurance, combining aerobic and resistance exercises to meet the specific needs of BCS. The dynamic mixed reality interface creates a more immersive, motivating experience, helping participants adhere to the program and achieve better rehabilitation outcomes.

The KE completed 8 weeks of exercise (total: 24 sessions), with three unsupervised 30-min sessions per week consistently conducted in a comfortable, safe environment with controlled temperature (22-23°C) and humidity (40%-50%). In week 1, the participants were closely supervised by trained professionals to effectively perform the KMR exercise program. Flexibility movements were incorporated into the warm-up and cool-down. The main exercise regimen comprised bodyweight resistance exercises targeting large and core muscles and aerobic and resistance exercises, with 40-second work and 20-second rest intervals (Table 1). This structure, based on HIIT, aims to maximize cardiovascular and muscular endurance by alternating periods of maximum effort with short rest breaks. The exercises are carefully selected to engage both upper and lower body muscles, ensuring a comprehensive workout that enhances strength, stability, and overall fitness, particularly tailored for the needs of BCS. The KE performed the scheduled KMR exercise program within the gymnasium and received real-time feedback (“Great,” “Good,” and “Bad”). Using mixed reality, and automatically saves data on adherence, including the movement repetitions in the server. The initial exercise was presented, with an audible count of repetitions; then, the next exercise was displayed on the screen with a 20-second timer, followed by a second exercise. This sequence was repeated for all designated exercises.

Table 1.

Exercise and stretching programs

The exercise intensity was personalized by tracking movement repetitions using automatically recorded data from the device. Each participant’s relative intensity was set based on performing exercises to maximum effort within a given time frame. This method ensured personalized and adaptive intensity, reflecting each participant’s maximal effort level. While indicators like heart rate and rating of perceived exertion can provide detailed insights, our approach focused on maximizing individual effort to account for variations in fatigue and effort. Participants could access their exercise records and performance evaluations of previous sessions. Every week, based on the participants’ performance history, the target number of repetitions for each exercise was incrementally increased by 5%-10%, and the adjusted targets were displayed on the screen, which enabled participants to gradually increase their exercise volume. Participants provided ongoing feedback to address concerns, preferences, and adapt the intervention as needed. Missed sessions were rescheduled after determining reasons, prioritizing adherence.

In contrast, the HS received videos of general health recommendations for stretching exercises (total: 24 sessions) (Fig. 2). Stretching exercises comprised movements aimed at improving flexibility in the major muscle groups, including the upper body and chest, and shoulder mobility. Each stretch was performed for 40 seconds, followed by a 20-second rest interval before the next stretch (Table 1). All participants underwent post-intervention assessments.

The KMR device automatically recorded data on adherence with the exercise program, including the total number of repetitions and performance grades (missed, bad, good, and great) of each movement. The correctness of the movements was determined by assessing the ratio of good and great grades to the total number of repetitions.

Results

1. General characteristics

Table 2 presents the general characteristics of the 77 BCS enrolled participants. On average, the participants were 49.4±7.5 years old, with a body weight of 57.2±11 kg. No significant intergroup differences (p > 0.05) in the descriptive and dependent variables were observed at baseline.

Table 2.

General characteristics of the study participants

2. Physical function

1) Shoulder ROM

The pre- and post-test results for shoulder ROM are presented in Table 3. Repeated-measures analysis of variance revealed a significant group-by-time interaction for flexion in the operated arm (p=0.009), flexion in the non-operated arm (p=0.004), abduction in the non-operated arm (p=0.033), and adduction in the operated arm (p=0.026). Further analysis revealed statistically significant effects over time for specific movements, such as flexion in the operated arm (p < 0.001), flexion in the non-operated arm (p < 0.001), extension in the operated arm (p < 0.05), abduction in the operated arm (p < 0.001), and abduction in the non-operated arm (p < 0.001).

Table 3.

Changes in shoulder range of motion from the pre-test to the post-test

Compared to HS, the KE showed a significant increase of 11.5° in flexion of the operated arm, improving from 154.3±12.5 to 165.8±11.2 between the pre- and post-tests (Fig. 3A). A significant group-by-time interaction (F=7.23, p=0.009) and a large effect size (Cohen’s d=0.93) further highlighted the distinct improvement patterns. Similarly, for flexion of the non-operated arm, the KE demonstrated a considerable improvement from a pre-test mean of 158.2±13.8 to a post-test mean of 162.5±12.2 (F=0.16, p=0.070), with significant group-by-time interaction (F=8.04, p=0.004) with a large effect size (Cohen’s d=0.84), indicating diverse responses to the interventions (Fig. 3B). For extension of the operated arm (Fig. 3C), the KE group improved from 46.5±9.2 to 52.2±7.4, while the HS group increased from 47.4±9.1 to 50.2±8.5. A significant time effect (F=4.85, p=0.021) was observed, but the group effect (F=0.38, p=0.384) and group-by-time interaction (F=1.05, p=0.310) were not significant. For extension of the non-operated arm (Fig. 3D), there was no statistically significant difference between groups.

Fig. 3.

Results of shoulder range of motion from pre- and post-test across the two groups. (A) Flexion operated arm. (B) Flexion non-operated arm. (C) Extension operated arm. (D) Extension non-operated arm. (E) Abduction operated arm. (F) Abduction non-operated arm. (G) Adduction operated arm. (H) Adduction non-operated arm. Two-way repeated-measures analysis of variance is used for statistical repeated ANOVA. ⁺p < 0.05 for group effect ^#p < 0.05 and ^###p < 0.001 for time effect; ^*p < 0.05 and ^**p < 0.01 for group-by-time interaction. HS, home stretching group; KE, Kinect-based mixed reality device exercise group; SD, standard deviation.

For abduction of the operated arm, the KE group showed a significant improvement from a pre-test mean of 150.2±23.9 to a post-test mean of 159.1±25.9. The analysis revealed a significant main effect of time (F=39.41, p < 0.001), while the group effect (F=0.06, p=0.049) and group-by-time interaction (F=0.96, p=0.683) indicated a moderately large effect size (Cohen’s d=0.74), demonstrating varied responses (Fig. 3E). For abduction of the non-operated arm (Fig. 3F), the KE group improved from 154.8±31.6 to 161.1±28.1, while the HS group increased from 157.0±23.0 to 163.3±27.8. The analysis showed a significant main effect of time (F=20.83, p < 0.001) and a significant group-by-time interaction (F=6.48, p=0.033), while the group effect was not significant (F=3.32, p=0.217).

As shown in Fig. 3G, for adduction of the operated arm, the KE group demonstrated a significant increase from a pre-test mean of 46.5±9.1 to a post-test mean of 52.6±7.2. The analysis revealed a significant group effect (F=0.59, p=0.023) and a significant group-by-time interaction (F=1.34, p=0.041), with a moderately large effect size (Cohen’s d=0.72), indicating different patterns of improvement over time between the groups. However, the time effect was not significant (F=3.16, p=0.228), suggesting that the changes were primarily driven by group differences. For adduction of the non-operated arm (Fig. 3H), there was no statistically significant difference between groups.

We conducted post hoc analyses on the variables that showed significant interaction effects (Table 4). Only the KE group demonstrated a significant mean difference in shoulder ROM variables, with notable changes observed in flexion of the operated arm (11.5±9.5), flexion of the non-operated arm (8.3±12.1), abduction of the operated arm (8.9±12.7), and adduction of the operated arm (6.1±10.4). In contrast, the HS group showed no significant mean differences in any of these four variables. Improvements in shoulder ROM were observed only in the KE, driven by the significant interaction effects.

Table 4.

Changes in shoulder range of motion by groups

2) HGS

Repeated-measures analyses of variance revealed no significant group-by-time interaction for HGS (F=2.97, p=0.075) (Fig. 4A). However, group comparisons revealed distinct values in the KE and HS. Specifically, the pre-test mean for the KE was 19.3±2.4, which was increased to 20.4±3.1 with a moderate effect size (Cohen’s d=0.52) in the post-test.

Fig. 4.

Results of hand grip strength (HGS) and aerobic capacity from pre-test and post-test across the two groups: HGS (A) and maximal oxygen uptake (VO₂max) (B). Two-way repeated-measures analysis of variance is used for statistical repeated ANOVA. ⁺p < 0.05 for group effect; ^#p < 0.05 for time effect. HS, home stretching group; KE, Kinect-based mixed reality device exercise group.

3) Aerobic capacity

Repeated-measures analysis of variance revealed a non-significant group-by-time interaction for VO₂max (p=0.357) (Fig. 4B), without significance, for both KE and HS.

3. Questionnaire

1) QOL (FACT-B)

The KE exhibited a significant increase from a pre-test mean of 61.31±9.88 to a post-test mean of 65.26±10.12, as did HS (pre-test mean: 60.25±10.29 increasing to 62.11±8.84 in the post-test) (Fig. 5A), without a significant main effect of group (F=0.23, p=0.634) or a significant group-by-time interaction (F=3.48, p=0.101).

Fig. 5.

Results of questionnaire from pre- and post-test across the two groups: Functional Assessment of Cancer Therapy-Breast (FACT-B) (A), Korean version of Disabilities of the Arm, Shoulder, and Hand (K-DASH) (B), and Korean version of International Physical Activity Questionnaire (K-IPAQ) (C). Two-way repeated-measures analysis of variance is used for statistical repeated ANOVA. ^##p < 0.01 for time effect; ^*p < 0.05 and ^**p < 0.01 for group-by-time interaction. HS, home stretching group; KE, Kinect-based mixed reality device exercise group; SD, standard deviation.

2) Shoulder disability (K-DASH)

The KE exhibited a significant reduction from a pre-test mean of 46.81±9.18 to a post-test mean of 40.87±9.31, as did the HS (pre-test mean: 43.85±10.69 increasing to 42.28±9.67 in the post-test) (Fig. 5B).

The statistical analysis underscored a significant group-by-time interaction (F=3.48, p=0.001) with a large effect size (Cohen’s d=0.84), indicating distinctive patterns of change over time between the KE and HS.

3) Physical activity (K-IPAQ)

Notably, the statistical analysis revealed a significant group-by-time interaction (F=7.18, p=0.011) with a moderately large effect size (Cohen’s d=0.73), highlighting discernible trends in the changes over time between the KE and HS (Fig. 5C).

4. Body composition

The results of %ECW with the exercise program for the analysis of body composition and edema are presented in Table 5. For %ECW, a significant main effect for the group (F=5.67, p=0.028) indicated a significant difference in %ECW between the KE and HS. The KE had a mean %ECW of 39.1±0.5, whereas the HS had a mean %ECW of 38.1±0.3. However, the main effects for time (F=0.061, p=0.937) and group-by-time interaction (F=0.006, p=0.918) were not significant.

Table 5.

Changes in body weight, BMI, and %ECW from the pre- to post-test

Discussion

BCS frequently experience a decline in physical function and QOL, in addition to increased fatigue. Exercise interventions are recommended to alleviate these negative effects [23-25]. We evaluated the effectiveness of an 8-week KMR exercise program on various measures of physical function, QOL, fatigue, and body composition in BCS. The main finding of our study was as follows: the 8-week KMR exercise program significantly improved the shoulder ROM, upper extremity function, and PA levels in BCS, highlighting the potential benefits of KMR as a novel exercise intervention.

Our results regarding the positive effects of exercise interventions on physical function and QOL in BCS align with the findings of Aydin et al. [26] and Basha et al. [13]. The improvements observed in our study, such as enhanced shoulder movements, reduced upper extremity disability, and increased PA levels, were consistent with the outcomes reported by Lang et al. [27], who reported an improvement of approximately 10° in both right and left flexion shoulder ROM, which was beneficial for BCS with difficulty in performing overhead work-related functional tasks. Fisher et al. [28] reported that a minimum shoulder flexion of 148° was necessary to reach a high shelf. However, the average flexion of 161° in both directions observed after the exercise intervention in this study exceeded this requirement. The exercise program in this study, which primarily included stretching exercises targeting the pectoralis major, anterior deltoid, and coracobrachialis muscles, and movements such as high knee chop and overhead sideband, may have influenced the results.

Recent studies have demonstrated that combined exercise interventions, including resistance training, aerobic activities, and dynamic stretching, offer superior benefits for improving ROM in BCS compared to static stretching alone [29,30]. These combined programs not only improve flexibility but also enhance muscle strength, neuromuscular adaptations, and joint mobility, leading to better overall functional capacity. This may be since resistance training activates muscles more effectively, while dynamic exercises improve circulation and reduce stiffness. Tailored exercise programs that address individual needs, pain levels, and physical conditions are likely to yield the best outcomes.

Basha et al. [13] demonstrated that VR-based exercise programs can effectively manage BCRL and improve QOL and physical functioning in BCS. While our study did not directly compare VR training with resistance exercises, our findings support the positive impact of exercise interventions on improving physical function and QOL in this population.

Collectively, these findings indicate that exercise interventions play a pivotal role in enhancing physical function and QOL in BCS. Our tailored exercise program utilizing KMR technology for BCS distinctly differed from those used in previous studies that utilized exergaming or VR interventions [14,31] and validated the significant effects of interventions, which used existing games. This distinction highlights the combination of a scientifically based customized exercise program and the advantages of KMR technology for BCS for targeted, effective intervention. The high adherence rate (92.4±8.1%) and improved performance rate suggest that the KMR exercise program is both feasible and acceptable for BCS. KMR’s real-time feedback and personalized progression maintained participant engagement, while the controlled environment contributed to strong adherence. These findings highlight the KMR’s potential for promoting long-term PA and rehabilitation.

An advantage of KMR over VR is the ability to recognize user movements and provide real-time feedback to guide correct exercise techniques. Although many studies have used VR interventions for BCS, most have focused on psychological distress [32] or effective distraction interventions for managing pain and anxiety [33], and frequently reported one-sided visual effects [34]. In our study, KMR technology assessed user movements in four stages, aiding in correct exercise techniques and posture, contributing to positive outcomes. The accuracy of exercise movements and ability to perform exercises in the correct posture significantly contributed to the positive outcomes of our study. The KMR provides interactive, personalized exercise guidance, ensuring safe and effective workouts for BCS, yielding positive outcomes.

As in the study by Yasunaga et al. [21], we evaluated the effects of exercise on edema by measuring the changes in %ECW in BCS as an indicator of BCRL. The group effect was the only significant finding in our study; however, the results are consistent with those reported by Yasunaga et al. [21] regarding the potential of %ECW as a valuable measure for assessing edema and lymphedema in BCS. Further research with larger samples is required to explore exercise interventions for %ECW and their potential in managing BCRL.

The results on aerobic capacity indicated no significant difference in the group-by-time interaction, which resemble the findings of Lee et al. [35]. They reported that the HIIT group maintained VO₂max, whereas control group experienced a significant decrease in VO₂max from baseline to 8 weeks. The decline in cardiorespiratory fitness may persist for years after cancer treatment [36]. The maintenance of aerobic capacity in our study constitutes a positive outcome.

Despite the absence of significant group-time interactions, the KMR exercise program showed positive effects on QOL in BCS, consistent with prior studies on exercise interventions [37]. Longer-duration studies (1-5 years), such as those of Odynets et al. [38] and Roine et al. [39], consistently reported QOL improvements, with FACT-B scores increasing from 82.5±1.9 to 119.2±3. Thus, QOL is a multifaceted concept that requires a relatively long time for meaningful changes, and the short-term intervention in this study may have influenced research outcomes.

Our analyses revealed a substantial interaction effect between group and time variables in relation to shoulder ROM, disability, and PA whereby the KMR exercise program effectively mitigated shoulder pain and enhanced PA in BCS.

The KMR exercise program is a promising intervention for enhancing physical function, QOL, and overall well-being among BCS. Significant improvements in various physical measures and the reduction in shoulder pain and fatigue highlight the potential benefits of customized exercise interventions. The interactive and personalized nature of the KMR contributes to the effectiveness and safety of exercise interventions, enabling individualized guidance and real-time feedback for enhanced engagement and motivation of BCS for improved outcomes. This study could serve as a catalyst for further research on exercise interventions for BCS.

While the average duration since surgery in our participants was 63 months, the key focus of this study lies in the use of a KMR exercise program. This intervention provides a unique approach to maintaining physical function and enhancing QOL. Previous studies [40,41] have explored exercise interventions during earlier stages of treatment, but the KMR system in this study offers long-term survivors an innovative and engaging method to improve physical function. The system integrates real-time feedback and personalized progression, making it a valuable tool for ongoing rehabilitation in survivors. By emphasizing the mixed reality component, this study highlights the innovative application of technology to address the long-term needs of cancer survivors, which may offer more comprehensive and sustained benefits compared to traditional exercise programs.

Article information Continued