2.1 Subjects
The study was conducted with the approval of the Ethics Committee of Zhejiang Normal University (Approval No. ZSRT2023003) and was registered with the China Clinical Trial Registry (Approval No. ChiCTR2300073580), and the first registration date is 2023/7/14. The recruited subjects volunteered to participate in the study, and all provided informed consent by signing the experimental informed consent form. Prior to the test, subjects were informed about the study’s content and requirements. They were also advised to adhere to certain guidelines: (1) avoid engaging in strenuous physical activities such as ball games, calisthenics, and hiking in the first three days of the test; (2) refrain from consuming coffee, alcohol, tea, and spicy foods in the initial three days of the test.
The study utilized G*Power sample size estimation software to determine the required sample size for the repeated measures ANOVA test. The selected parameters included an F-test, ANOVA: Repeated Measures Within-Between Interaction, effect size of 0.125, alpha error of 0.05, and power of 0.80. This analysis determined that 33 participants were needed for each of the four groups, resulting in a total of 132 participants to ensure statistical validity. To account potential sample loss, 140 university students (72 males and 68 females) were initially recruited. The participants were randomly assigned to the V-M group, M-V group, V group, and M group. However, due to equipment failure and issues with data reception during the experiment, the final number of valid subjects included in the analysis was 132, with 68 boys and 64 girls. The basic characteristics of the subjects are detailed in Table 1.
Table 1
Basic characteristics of study subjects.
| V-M Group | M-V Group | M Group | V Group | P-Value |
n | 33 | 33 | 33 | 33 | |
Age | 24.09 ± 0.81 | 24.09 ± 0.91 | 24.09 ± 1.10 | 23.82 ± 0.73 | 0.423 |
Height /cm | Male | 176.76 ± 4.52 | 177.71 ± 8.45 | 175.78 ± 4.81 | 173.76 ± 5.19 | 0.517 |
Female | 165.31 ± 3.57 | 164.38 ± 5.07 | 164.80 ± 5.48 | 165.44 ± 4.56 | 0.257 |
Weight /Kg | Male | 71.29 ± 9.08 | 76.82 ± 6.40 | 71.89 ± 7.91 | 70.82 ± 7.66 | 0.267 |
Female | 55.75 ± 2.82 | 55.69 ± 5.98 | 54.33 ± 9.21 | 54.31 ± 3.89 | 0.986 |
BMI /(kg·m− 2) | Male | 21.71 ± 2.10 | 22.55 ± 2.68 | 21.76 ± 3.03 | 21.68 ± 2.44 | 0.821 |
Female | 21.47 ± 3.10 | 21.77 ± 2.45 | 21.30 ± 2.23 | 21.57 ± 2.30 | 0.355 |
An independent team member, not involved in other stages of the research project, was responsible for randomly assigning participants. Each recruited participant received a code, and after the baseline data collection, they were randomly allocated to either the experimental or control conditions. The individuals conducting the randomization were unaware of the participant’s circumstances, ensuring a blind assignment process. Furthermore, the data collectors remained unaware of the participants’ groupings throughout the study period. The subject assignment process is illustrated in Fig. 1.
2.2 Music and VR programs
In this study, two types of music were chosen for the music intervention based on their tempo. The selection process involved using the Mix Meister BPM Analyzer software to determine the number of beats per minute (bpm) of the music. Following the tempo classification by Lee and Kimmerly (2014), music with a fast tempo ranging from 120 to 150 bpm and music with a slow tempo ranging from 55 to 90 bpm were selected(Lee and Kimmerly, 2014). The fast-tempo music was used during the music intervention phase while the subjects were engaged in the MONARK power cycling exercise from Sweden. On the other hand, slow-tempo music was employed during the music intervention phase when the subjects were in the post-exercise recovery period. Given the individual differences in the subjects’ music preferences, the subjects could choose the music according to their needs. To align with the public’s preferences for fast and slow-tempo music, network data analysis was conducted to select music tracks with high audience appeal (shown in Table 2).
During the music intervention in the experiment, subjects were instructed to wear sports noise-canceling headphones, specifically the Beats Solo3 Wireless model. This measure aimed to reduce any potential interference in the experimental results caused by the noise generated by the power car. The volume of the headphones was set to a maximum of 75 dB, and subjects had the flexibility to adjust the volume within a range of 10 dB according to their preference. The VR intervention in the experiment involved subjects immersing themselves in natural environment videos while wearing Pico Neo VR glasses all-in-one. The selected video, sourced from the Internet, featured a green natural environment, and had a duration of at least 15 minutes. Throughout the video playback, the audio was muted. The content of the video showcased various elements of natural green vegetation such as green grass, trees, shrubs, jungles, rivers, waterfalls, and more, as illustrated in Fig. 2.
Table 2
Music | Duration | Rhythm(beats/min) |
Catch My Breath | 4min10s | 125(fast) |
Remember Our Summer | 2min43s | 128 |
Stronger (What Doesn’t Kill You) | 3min41s | 130 |
Lost in the Discotheque (Radio Edit) | 3min31s | 143 |
River Flows In You (Original Mix) | 4min58s | 128 |
Wake | 4min31s | 130 |
I Love You | 4min22s | 69(slow) |
To Me | 4min17s | 78 |
Silver City | 3min52s | 75 |
Love Is Gone | 2min56s | 70 |
Critical | 3min10s | 83 |
So Far Away | 2min51s | 74 |
2.3 Experimental site and time
This study was carried out in the Exercise Science Laboratory of Zhejiang Normal University, located in China. The laboratory’s environment and soundproofing were deemed satisfactory after inspection and comparison, contributing to the smooth execution of the experiment. The research was conducted between September and November 2023, with testing taking place on the same day from Monday to Friday, spanning the hours of 8:00 a.m. to 11:00 a.m. and 3:00 p.m. to 5:00 p.m.
The study site’s climate during the experimental period was favorable, with minimal temperature variation from the beginning to the end of the experiment. This stability helped reduce the impact of natural environmental factors such as air temperature and humidity on the study outcomes.
2.4 Experimental process
(1) Preparation stage: The experimenter ensured that the relevant experimental equipment was set up for each test group before the subjects arrived at the experimental test site. Upon the subjects’ arrival, the experimenter provided an introduction to the test instructions, briefing them on the procedures before commencing the formal testing. This preparation phase typically lasted approximately 3 minutes, allowing the subjects to familiarize themselves with the upcoming tasks and ensuring a smooth transition into the formal testing phase.
(2) Pre-test stage: The subjects’ individual demographic and social variables, daily physical activity level, and other relevant information were collected. Following this, the dependent variables were collected in the following order: ①Measurement of the systolic and diastolic blood pressure of the subjects using an Omron electronic sphygmomanometer; ②Collection of heart rate variability data using the First Beat wearable wireless physiological device, specifically 5 minutes before the experiment; ③Completion of real-time assessments using the “Positive Affect Scale” and “Negative Affect Scale” by the subjects. This comprehensive data collection process lasted approximately 8 minutes and aimed to establish baseline measurements and assess the subjects’ physiological and affective states before the formal testing procedures commenced.
(3) Intervention stage: Subjects were directed to wear either Pico Neo VR glasses or sports noise-canceling headphones with the volume set at 75 ± 5 noise level, ensuring that the ambient sound level in the laboratory remained below 40 dB. The subjects commenced a 15-minute session of moderate-intensity aerobic power cycling. The exercise load was adjusted to 60–69% of each individual’s maximal heart rate, with heart rate monitoring to maintain a range of 120–150 beats/min. Following the aerobic cycling phase, subjects proceeded with 2-minute power cycling intervals at an intensity of 20%-30% of their maximal heart rate. This interval period allowed the subjects’ heart rates to gradually return to their resting rate without any VR or music interference. Subsequently, the subjects were instructed to maintain a sedentary position while continuing to wear VR glasses or sports noise-canceling headphones for an additional 15 minutes. The specific experimental flow chart detailing these instructions and activities can be found in Fig. 3, illustrating the sequence of events during the experimental procedure.
V-M group: Subjects watched a natural environment video during exercise and experienced music intervention after exercise. This group engaged in 15 minutes of aerobic power cycling with a natural environment video, followed by 2 minutes of interval exercise, and concluded with 15 minutes of sedentary rest accompanied by slow music. M-V group: Subjects listened to music during exercise and viewed a natural environment video after exercise. They participated in 15 minutes of aerobic power cycling with fast-paced music, followed by 2 minutes of interval exercise, and ended with 15 minutes of sedentary rest combined with a video of the natural environment. M group: This group received music interventions during and after exercise. They performed 15 minutes of aerobic power cycling with fast-paced music, followed by 2 minutes of interval exercise, and concluded with 15 minutes of sedentary rest accompanied by slow-paced music. V Group: Subjects watched videos of the natural environment during and after exercise. They engaged in a 15-minute aerobic power cycling session with natural environment videos, followed by a 2-minute interval session, and ended with a 15-minute meditation break with videos of the natural environment.
(4) Post-test stage: At the end of the intervention, the heart rate belt was kept on until the subject’s heart rate had returned to the baseline quiet state prior to the start of the experiment. This return to the quiet state was maintained for 30 seconds or longer compared to the heart rate prior to the experiment initiation. The post-test phase included the following assessments:①Systolic and diastolic blood pressure measurements were taken using an Omron electronic sphygmomanometer; ②Heart rate variability data was collected using the First Beat wearable wireless physiological device, specifically 5 minutes before the conclusion of the experiment; ③Subjects were required to complete real-time assessments using the “Positive Affect Scale” and “Negative Affect Scale” to gauge their affective states following the intervention. These post-test measures aimed to evaluate the physiological and affective responses of the subjects after the completion of the intervention. The photos of the experiment site are shown in Fig. 4.
2.5 Dependent variable
(1) Heart rate variability
In the controlled combined exercise–music intervention experiments, the low frequency to high frequency (LF/HF) ratio is often used as the frequency domain analysis indicator. The LF/HF ratio can determine the equilibrium of sympathetic and vagal nerves or the modulation degree of the sympathetic nerves. The time domain analysis indicator primarily uses the root mean square of the difference between adjacent full RR intervals (RMSSD) and the standard deviation of continuous regular RR intervals (SDNN) as comprehensive reflective markers of the effect of short-intervention control experiment on affect improvement(Koelsch, 2018; Jacquet et al., 2021). Therefore, LF/HF, RMSSD, and SDNN data were collected in this experimental study as the metrics for processing and analysis.
For the acquisition of the above heart rate variability (HRV) indices, the First Beat Sports wireless physiological data collection system with the ECG module device was selected for the experiment. The apparatus can detect and capture changes in the subject’s heart rate in real-time and during the activity. Simultaneously, the changing signal of the subject’s heart rate can be automatically converted into time- and frequency-domain data in the background for recording and storage. The selected device has been used in many controlled experiments involving green fitness and gardening activities. The real-time accuracy and reliability of the data recorded by the aforementioned device system have been verified in many controlled experiments (Light et al., 2012; Beauchaine and Thayer, 2015)
(2) BP
Blood pressure includes systolic blood pressure (SBP) and diastolic blood pressure (DBP), the reduction of blood pressure to a certain extent can reflect the reduction of the negative effect on the subjects, and the improvement effect of the blood pressure indicator can be reflected from the side of the human body to improve the effect of short-term affect (Pretty, 2004; Zijlema et al., 2018). In this experimental study, the systolic and diastolic blood pressure of the subjects will be measured using an Omron upper arm electronic sphygmomanometer. After one measurement is completed and the readings are stabilized, the data will be recorded and the procedure will be repeated after 1–2 minutes, and the final value will be the average of the two measurements. To avoid the possibility of differences in blood pressure measurements between the pre-test and post-test due to different arms, the arms of the subjects should be kept in the same position during the pre-test and post-test blood pressure measurements.
(3) Short-term affect
The "affect" indicator can be a good way to record and assess the real-time mental health status of the subjects promptly. For the real-time affect condition of the subjects, the experiment used the International Positive/Negative Affect Scale (short version), which has been proven to be widely used and reliable in many controlled experiments(Thompson, 2007; Liu et al., 2012). During the pre-test and post-test phases, the subjects will rate the positive and negative questions according to their real-time affective state. Finally, the affective state of the subjects in the pre-test and post-test phases will be reflected by calculating the total score values of the positive and negative scores respectively.
2.6 Statistical methods
(1)The collected data were initially entered into Excel 2010 for storage and later imported into SPSS 22.0 software for statistical analysis. Descriptive statistics were utilized to analyze the demographic and sociological variables such as gender and age within the sample, presenting the results in terms of percentages or mean standard deviation. To compare the effect of improvement between groups in the pre-test and post-test phases, a multifactor ANOVA was conducted. The analysis considered the F-value, p-value, and effect size (ɳp2) to assess the significance and magnitude of the observed differences. Furthermore, differences in the effects of improvement between groups were evaluated using repeated measures ANOVA. This analysis also took into account the F-values, p-values, and effect sizes (ɳp2) to determine the significance and impact of these differences over time. Each statistical model was adjusted for various covariates including gender, age, height, weight, and BMI to control for potential confounding variables and ensure a more accurate assessment of the impact of the interventions on affect improvement.
(2)The HRV data were imported into an Excel sheet to process calculations, and the export was divided into a time domain analysis table and a frequency domain analysis table. The calculations used were all corrected RR values, and the specific calculations for RMSSD, SDNN, and LF/HF were as follows:
Time domain indicators
1) SDNN
Defined as: the standard deviation of the continuous normal RR period.
Excel calculations were made by applying the standard deviation formula to the RR values. Take cell A1:A4 data as an example: =STDEVP (A1:A4).
2) RMSSD
Defined as: the root mean square of the difference between adjacent RR intervals.
Excel calculations were as follows: (1) The RR interval value column was copied and then staggered before and after the “—” in front of the value, and then “+” dropped down, resulting in the adjacent interval difference. (2) The root mean square was then calculated. Cell A1:A4 data were used as an example:
= SQRT (SUMSQ (A1:A4)/ COUNTA (A1:A4)).
Note
SQRT calculates the square root of an arithmetic number; SUMSQ calculates the sum of squares of several numbers; and COUNTA calculates the total number of numbers.
Frequency Domain Indicators
3) LF/HF
Defined as: the ratio of LF to HF.
Excel calculates it as = LF /HF, then “+” drops down to obtain the average value of LF/HF for that time duration.