Participants
This study enrolled 71 healthy participants (29 males and 42 females; mean age: 22.3 ± 2.2 years; mean weight: 166.6 ± 8.8 kg) who did not experience pain while cycling. We excluded pregnant individuals as well as those with vestibular, neurological, cardiopulmonary, psychological, or musculoskeletal disorders. The participants provided written informed consent. This study approved by the University Institutional Review Board for Human Investigations (number: jjIRB-211026-HR-2021-1028). A statement to confirm that all methods were carried out in accordance with relevant guidelines and regulations.
Sahrmann Core Stability Test (SCST)
The SCST was performed to evaluate core stability. It included five progressively more difficult tasks. The inflatable pad of a stabilizer pressure biofeedback unit (Chattanooga Group, Hixson, TN, USA) was placed in a natural lordotic curve while participants were placed in a crooked lying position. The pad was inflated to 40 mmHg before the task. A deviation of >10 mmHg during the task indicated loss of stabilization of the lumbopelvic hip complex by the stabilizer muscles. Participants who completed a task without a deviation of >10 mmHg were instructed to perform the next task. Performance (i.e., the ability to complete the tasks without a deviation of >10 mmHg) was rated on a 5-point scale. The participants were divided into poor and good-core-stability groups based on their SCST scores (0–1 and 2–5, respectively). The tasks were performed as reported previously28.
Instruments
A high-resolution single IMU (BNO080; Ceva Technologies, Rockville, MD, USA) equipped with a triaxial accelerometer and triaxial gyroscope was embedded into a left-side wireless earbud (QCY-T1C; Dongguan Hele Electronics, Dongguan, China) to measure head angle (Fig. 1). IMU data were collected at 100 Hz. Each sample contained signed 16-bit acceleration output for x, y, and z axes. The acceleration outputs were transferred to a self-developed mobile app via Bluetooth. The app calculated the mediolateral head angle in real time, similar to a previous study29, and controlled the volume of music from the wireless earbud to provide feedback.
Feedback music
Feedback music was provided in real time to prevent excessive mediolateral head motion during cycling. If the mediolateral head angle exceeded a predefined threshold, the wireless earbud on the side of the head tilt was muted. Once the mediolateral angle returned to the set range, the muted earbud was unmuted (Fig. 2). For example, if the angle threshold was 10° and the head tilted >10° to the right side, the earbud on the right side was muted; the muted earbud was unmuted when the mediolateral head angle was reduced to <10°.
Cycling
Participants wore the wireless earbud, including the IMU sensor, in their ears to measure head angle in the frontal plane and receive feedback music. During warm-up, participants cycled for 5 min at their preferred speed. Then, after a 5 min rest period, the participants were instructed to cycle at the fastest speed possible; 70% of the measured maximum speed was set as the target speed.
The participants cycled at the target speed with and without feedback music. In the trial with no feedback, participants were asked to cycle at the target speed for 1 min. The mediolateral head angle was measured for 1 min during cycling at the target speed, and data from the final 40 s were analyzed. A value of 50% of the measured maximum mediolateral head angle was set as the threshold for feedback music. After the trial with no feedback, participants cycled for 1 min at the target speed and received feedback. The participants rested for 3 min between the trials with and without feedback.
Symmetry
The maximal right and left head angles during each cycle were used to evaluate the range of mediolateral head motion and symmetry in head angle during cycling. Positive and negative signs for head angle represent the right and left directions, respectively. Therefore, the maximal value was the maximal right head angle, and the minimal value was the maximal left head angle (Fig. 3A).
The maximum and minimum values during each cycling cycle were recorded. The average difference between the maximum and minimum values indicated the range of mediolateral head motion:
$$amp=\frac{\sum _{i}({max}_{i}–{min}_{i})}{n}$$
Here n is the number of peak values and max and min represent the maximum and minimum values during each cycling cycle, respectively (Fig. 3B).
Head angle symmetry during cycling was represented by the average of the maximal and minimum values:
$$SI= \frac{\sum _{i}\left|({max}_{i}+{min}_{i})/2\right|}{n}$$
For a perfectly symmetric mediolateral head angle, the maximal right and left head angles should be equal (i.e., symmetry index [SI] = 0). Because the values for each direction have opposite signs, the closer the SI value is to 0, the more symmetric the head angle is in the frontal plane (Fig. 3B).
Statistical analyses
The independent variables included core stability and feedback music, whereas the dependent variables included amp and SI. A 2 × 2 analysis of variance (ANOVA; core stability group × feedback) with a mixed-model design was performed to test the effects of the independent variables on the dependent variables. Differences were considered significant when p < 0.05. Paired post hoc t tests were performed to evaluate differences in the dependent variables between the trials with and without feedback music in each group. An independent t test was performed to compare the dependent variables between the good and poor-core-stability groups in the trial without feedback music.