Physical Therapy Korea

The serratus anterior (SA) muscle, described as the primary mover and stabilizer of the scapula, can produce scapular protraction, upward rotation, and posterior tilt while stabilizing the medial border and inferior of the scapula to the thoracic wall [1,2]. Weakness of the SA muscle can lead to overactivation of the pectoralis major (PM) and upper trapezius (UT) muscles during shoulder elevation, which can lead to upper extremity dysfunction, including scapular winging and subacromial impingement [2-4]. Therefore, previous therapeutic exercise program for SA activation for the rehabilitation and prevention of scapular dysfunction have been investigated [2].

The push-up plus (PP) exercise generated the highest SA muscle activation as compared with other scapular stabilization exercises and was recommended for high activate the SA muscles [5-7]. However, it is necessary to control exercise intensity because PP is difficult to perform repetitively in patients with weakened SA muscles. Therefore, modifications to the PP, including knee PP (KPP), elbow PP, and wall PP, are recommended for individuals with scapula winging or at an early stage of rehabilitation [8]. These researchers reported that SA activity was lower but still relatively high during KPP compared to conventional PP [8].

The PM muscle, which has prime role of shoulder horizontal adduction, was activated with the SA muscle during scapular protraction exercises. During PP exercises, individuals with scapular winging have higher PM activity than healthy individuals [3,4,8]. Since a lower SA/PM ratio may contribute to shoulder joint dysfunction, an imbalanced SA and PM ratio can be adverse for individuals with scapular winging [4]. Prior investigators reported that when isometric horizontal abduction (IHA) was applied during PP, PM muscle activity decreased because of reciprocal inhibition, while SA muscle activity increased [4]. Therefore, IHA can effectively increase SA muscle activation while preventing the overactivation of PM muscle [4,9].

A previous study found that electromyography (EMG) activity of the SA muscle was significantly higher during shoulder exercise between 120° and 150° of shoulder flexion [5]. It was also found that the PP in a humeral position of 120° of shoulder flexion was greater activation in SA activation than conventional PP, which is at 90° of shoulder flexion, in healthy individuals [10]. Researchers have explained that the activity of the SA muscle was high at 120° of shoulder flexion because its main function was upward rotation and protraction of the scapula in the thoracic wall [10].

To the best of our knowledge, to date, no study has compared the effect of IHA application on SA, PM, and UT muscle activities at two different shoulder flexion angles during KPP in individuals with scapular winging. Accordingly, this study aimed to compare the effects of IHA and different shoulder flexion angles (90° and 120°) on SA, PM, and UT muscle activity and on SA/PM and SA/UT muscle activity ratios during KPP in individuals with scapular winging. We hypothesized that IHA application and 120° of shoulder flexion during KPP can increase the activity of the SA and UT muscles, decrease the activity of the PM muscle, and increase the ratio of SA/PM and SA/UT activity.

1. Participants

A power analysis was performed using G-power software version 3.1.2 (Franz Faul, University of Kiel). Data from a six-person pilot study were used to calculate the sample required to achieve a power of 0.80, an effect size of 0.44, and α level of 0.05. A total of 16 participants were necessary. Participants were recruited based on the presence of scapular winging, as determined by scapulometer measurement [11]. Participants were included if the distance between the inferior angle of the scapula and thoracic wall was > 2 cm [11]. Our study included 20 men with scapular winging, with an age of 24.65 ± 2.65 (range, 20–30) years, weight of 78.52 ± 13.26 kg, height of 175.57 ± 7.03 cm, body mass index of 25.39 ± 2.65 kg/m2, and a scapular winging of 2.74 ± 0.64 cm. Participants with past or current dysfunction or pain that substantially limited the range of motion or instability of the shoulder joint, adhesive capsulitis of the shoulder joint, or symptoms of cervical pain were excluded [4]. Participants who were unable to perform KPP during the familiarization session or who complained of discomfort or pain during measurement were also excluded. Prior to the study, all participants were thoroughly explained about the experimental process by the principal investigator and signed an informed consent. This study was approved by the Institutional Review Board of the Yonsei University Mirae campus (IRB no. 1041849-202107-BM-094-01).

2. Surface EMG Data Collection

Surface EMG (Noraxon TeleMyo DTS Wireless System; Noraxon Inc.) was used to measure SA, PM, and UT muscle activity. The sampling rate of the EMG signal was set to 1,500 Hz. The raw signal was filtered using a digital band-pass filter at 20–450 Hz. The measured data were analyzed using Noraxon MyoResearch XP software version 1.08 (Noraxon Inc.). Disposable Ag/AgCl surface electrodes were placed approximately 20 mm apart in the same direction of each muscle fibers. The attachment site of the surface electrodes were as follows: the electrodes of the SA muscle were attached directly below the axillary area at the same level of the inferior angle of scapula [12]; electrodes of the PM muscle were attached to the chest approximately 20 mm horizontally from the axillary fold [12]; and electrodes of the UT muscle were placed at the midpoint of the shoulder ridge between the acromion and the seventh cervical vertebrae [12]. The maximum voluntary isometric contraction (MVIC) of the SA, PM, and UT muscles were measured based on standard manual muscle testing positions, as recommended by prior investigators [13]. To measure the MVIC of the SA muscle, participants sat upright in a chair without support. The participant’s shoulder joint was elevated at 125° in the scapular plane and the investigator applied resistance to the participant’s elbow. The MVIC of the PM muscle was measured while the participants adducted the shoulder at the 90° of shoulder flexion against the investigator’s resistance in the supine position. To measure the MVIC of the UT muscle, participants were asked to maintain isometric abduction against the investigator’s resistance, with their arms elevated at 90°. The EMG data of each muscle was measured for 5 seconds and the mean MVIC amplitude was determined using only the value for the middle 3 seconds. Participants performed two rounds of each test, with 2-minute rest between each measurement to prevent the fatigue. The measured EMG values of the SA, PM, and UT muscles during the four KPPs were described as the percentages of the mean maximal voluntary isometric contraction (%MVIC).

3. Procedures

Prior to measurements, the principal investigator demonstrated how KPP is performed using two shoulder flexion angles with or without IHA for approximately 30 minutes to familiarize the participants with the process. The participants were not provided with the intent and results of this study. The order of measurement of KPPs was randomly assigned using Microsoft Excel (Microsoft Corp.). The participants performed three rounds of each KPP for measurement of SA, PM, and UT muscle activity. To minimize fatigue and test effects, participants were allowed a 5-minute rest between each KPP. The measured EMG values of the SA, PM, and UT muscles in four types of KPP were recorded as %MVIC.

The starting posture was supported by the knees and both the hands. The investigator determined that the width between each knee should be equal to the pelvic width, and the width between the hands should be equal to the width between the shoulders. The elbows were fully extended, the fingers were extended, the contacted thenar and hypothenar areas were equally loaded, and the shoulder and hip joints had an angle of 90°, as set by a goniometer. For data collection, participants were asked to protract their scapula as far as possible and maintain this position for 5 seconds. While maintaining the KPP posture, the alignment of the pelvis and spine was kept neutral. To maintain a constant scapular protracted posture during measurement, the participants were asked to keep their thoracic spine in contact with the target bar (Figure 1A).

Figure 1. KPP with or without IHA at 90° and 120° of shoulder flexion. (A) KPP90, (B) KPP120, (C) KPP90-IHA, and (D) KPP120-IHA. KPP, knee push-up plus; IHA, isometric horizontal abduction.

The starting position of KPP120 was the same as that of KPP90, but the shoulder flexion was 120°, as set by a goniometer. As with KPP90, participants protracted their scapula as much as possible and held this position for 5 seconds. Using the target bar set in the familiarization session, participants were asked to maintain the correct KPP120 posture and were provided feedback from researchers to maintain alignment of posture accurately (Figure 1B).

The starting posture of KPP with IHA at 90° of the shoulder joint was the same as that of KPP90, but the elastic band was used on the elbow to provide resistance to the shoulder horizontal abductors. To provide relative resistance for participants, the elastic band intensity was set to allow the shoulder horizontal abduction to be performed 10 times or less [8]. In KPP90-IHA, participants protracted and maintained the scapula using the target bar in the same manner as in KPP90. Moreover, the participants had both their elbow joints in contact with the target bar to maintain their upper arm posture against the resistance of IHA (Figure 1C).

The measurement posture was set by the investigator under the same conditions as that of KPP120. However, in KPP120-IHA, an elastic band was applied to the elbow joint to provide resistance to IHA of shoulder. As with KPP90-IHA, the length of the elastic band was set to the same resistance, and the target bars were used on the thoracic spine and both elbow joints to maintain the correct posture of the participant during the measurement (Figure 1D).

4. Statistical Analysis

The Kolmogorov–Smirnov test was used to confirm the assumption of normal distribution. Two-way repeated analyses of variance with two within-participant factors (exercise type: with and without IHA; shoulder flexion angle: 90° and 120°) were used to determine the statistical significance of SA, PM, and UT muscle activity and SA/PM and SA/UT muscle activity ratios. A p-value < 0.05 was considered statistically significant, and if a significant interaction of shoulder exercise type × flexion angle was found, a pairwise comparison with Bonferroni correction was used to determine the simple effect (α = 0.0125). Main effects of exercise type and shoulder flexion angle were determined when two-way repeated analyses of variance were revealed no significant interaction of exercise type × shoulder flexion angle. IBM SPSS version 20.0 (IBM Co.) was used for statistical analysis.

There was a significant interaction of exercise type × shoulder flexion angle in the muscle activity of the SA (F1,19 = 4.514; p = 0.047) and PM (F1,19 = 12.005; p = 0.003) muscles. There was no significant interaction for UT muscle activity (F1,19 = 0.543; p = 0.470), and the main effects of exercise type (F1,19 = 16.877; p = 0.001) and shoulder flexion angle (F1,19 = 16.877; p = 0.001) were significant. There was no significant difference in SA muscle activity between KPPs (Figure 2). PM muscle activity was significantly lower in KPP90-IHA than in KPP90 (p = 0.007; 95% confidence interval [CI] = 1.61–9.15; effect size [ES] = 0.92) and in KPP120 than in KPP90 (p = 0.006; 95% CI = 1.77–9.47; ES = 0.94; Figure 3). UT muscle activity was significantly higher with IHA than without IHA (%MVIC with IHA = 8.39%; %MVIC without IHA = 5.89%; Figure 4A) and at 120° of shoulder flexion than at 90° of shoulder flexion (%MVIC at 120° of shoulder flexion = 8.51%; %MVIC at 90° of shoulder flexion = 5.77%; Figure 4B). The SA/PM muscle activity ratio was significantly greater in KPP90-IHA than in KPP90 (p = 0.004; 95% CI = –4.66 to –0.93; ES = 1.24) and in KPP120 than in KPP90 (p = 0.001; 95% CI = –4.96 to –1.30; ES = 1.39; Figure 5). The SA/UT muscle activity ratio was a significant main effect of the exercise type (F1,19 = 14.111; p = 0.001), but the shoulder flexion angle did not have any significant main effect (F1,19 = 1.045; p = 0.319). The SA/UT activity ratio was significantly higher without IHA than with IHA (without IHA = 7.04; with IHA = 5.24; Figure 6).

Figure 2. Comparison of SA activity during KPP with and without IHA at 90° and 120° of shoulder flexion. SA, serratus anterior; %MVIC, percentages of the mean maximal voluntary isometric contraction; KPP, knee push-up plus; IHA, isometric horizontal abduction.

Figure 3. Comparison of PM activity during KPP with and without IHA at 90° and 120° degree of shoulder flexion. PM, pectoralis major; %MVIC, percentages of the mean maximal voluntary isometric contraction; KPP, knee push-up plus; IHA, isometric horizontal abduction. *p < 0.0125, significant simple effect.

Figure 4. Comparison of UT activity during KPP with and without IHA at 90° and 120° of shoulder flexion. UT, upper trapezius; %MVIC, percentages of the mean maximal voluntary isometric contraction; KPP, knee push-up plus; IHA, isometric horizontal abduction. (A) Significant difference in the main effect for IHA condition, (B) significant difference in the main effect for shoulder flexion angle. *p < 0.05, significant difference in main effects.

Figure 5. Comparison of the SA/PM muscle activity ratio during KPP with and without IHA at 90° and 120° of shoulder flexion. SA, serratus anterior; PM, pectoralis major; KPP, knee push-up plus; IHA, isometric horizontal abduction. *p < 0.0125, significant simple effect.

Figure 6. Comparison of the SA/UT activity ratio during KPP with and without IHA at 90° and 120° of shoulder flexion. SA, serratus anterior; UT, upper trapezius; KPP, knee push-up plus; IHA, isometric horizontal abduction. *p < 0.05, significant difference in the main effect for IHA.

We compared SA, PM, and UT muscle activity and the SA/PM and SA/UT muscle activity ratios according to the application of IHA and different shoulder flexion angles (90° and 120°) during KPP. Significant interactions between exercise type and shoulder flexion angle were observed in SA and PM muscle activity and the SA/PM muscle activity ratio. Moreover, UT activity was significantly greater with IHA than without IHA and in the 120° of shoulder flexion than at 90° of shoulder flexion. The SA/UT muscle activity ratio was significantly greater without IHA than with IHA.

PM activity was significantly lower in KPP90-IHA than in KPP90 and in KPP120 than in KPP90. These results support our research hypothesis. Previous researchers reported that there was a significant decrease in PM when IHA was applied during PP exercises and explained that the application of IHA was reciprocally inhibited by PM, which plays a role in horizontal adduction [9]. In our study, PM activity was significantly lower with IHA than without IHA in KPP90, but there was no significant difference when IHA was used in KPP120. Regardless of whether IHA was used, PM activity was significantly lower in KPP120 than in KPP90, and there was no additional effect of IHA in KPP120. The results of PM activity according to the shoulder flexion angle during KPP were inconsistent compared with those of previous studies [10]. In the previous study, there was no significant difference in PM muscle activity between shoulder flexions of 120° and 90° during PP in healthy individuals [10], but in our study, PM activity was significantly lower in KPP120 than in KPP90. The difference between healthy individuals in the previous study and individuals with scapular winging in our study seems to have affected the inconsistency between the results of the two studies. According to previous studies, during PP exercises, individuals with scapular winging increased the compensatory muscle activity of the PM muscle instead of that of the weakened SA muscle. Accordingly, PM muscle activity was significantly greater in individuals with scapular winging than in those without [14]. The results of our study demonstrated that, even though individuals with scapular winging displayed compensatory overactivity of PM in KPP90 [14], it was possible to significantly reduce the muscle activity of PM in KPP120. The main function of the PM muscle is horizontal adduction, but it also plays a supporting role in shoulder flexion up to an angle of 60° [15]. However, PM cannot move forward or upward at a shoulder flexion of 90° or more. It acts as an extensor and subsequently returns to the anatomical position [15]. Therefore, the compensatory overactivation of PM may be less intense in individuals with scapular winging with 120° of shoulder flexion (KPP120) than in those with 90° of shoulder flexion (KPP90).

Our results demonstrated significantly greater UT muscle activity during KPP with IHA than without IHA and 120°of shoulder flexion than 90° of shoulder flexion.; this result supports our research hypothesis. Clinically, the UT muscle is an overused compensatory muscle for stabilizing the muscles of the scapula in patients with shoulder pain [2]. An overactivated UT muscle may cause an abnormal upward rotation along with an excess elevation of the scapula, which may cause impingement of the shoulder joint [16]. Therefore, previous researchers studied the methods to minimize UT muscle activity while increasing the activity of stabilizer muscles of the scapula, such as the lower fiber of the trapezius and SA muscle [2,17]. A previous study compared the muscle activity of SA and UT among PP exercises in various postures, including PP, elbow PP, wall PP, and KPP. The researchers have reported that PP yielded the highest SA muscle activity and SA/UT muscle activity ratio, though this training was difficult to perform by patients; hence, elbow PP and KPP with relatively high SA muscle activity and SA/UT muscle activity ratios were recommended as alternatives [2]. Although the UT muscle is the primary muscle of scapular elevation [18], its muscle activity significantly increased with IHA in the scapular protraction state during KPP90. To maintain the IHA state in scapular protraction during KPP, scapular stability may be required more than when IHA is not used; accordingly, scapular muscle activity, including that of the trapezius, may be increased [18]. In a previous study, by comparing muscle activity according to shoulder flexion angle during conventional PP in healthy people, UT muscle activity was found to have significantly increased at 120° of shoulder flexion than at 90° of shoulder flexion, but there was no significant difference in the ratio of SA/UT activity between shoulder flexion angles [10]. Despite the differences in participants between the previous study and ours, the results of both studies showed that the UT muscle activity was significantly greater in the 120° shoulder flexion posture than in the 90° shoulder flexion posture [10]. Therefore, the reason for the higher activity of the UT muscle at 120° of shoulder flexion than at 90° of shoulder flexion can be interpreted as the difference in exercise posture, rather than the compensatory overuse for the weakened SA muscle. A previous researcher reported that, as the function of UT can change from the scapular supporter to the scapular upward rotator as the shoulder flexion angle increases [19], UT muscle activity can increase for upward rotation and elevation of the scapula during PP at 120° of shoulder flexion [10]. Therefore, an increase in UT muscle activity at 120° of shoulder flexion is essential and will not be considered unnatural.

The significant difference in the ratio of SA/PM activity was consistent with the muscle activity of PM. The ratio of SA/PM activity was significantly greater in KPP90-IHA than in KPP90 (KPP90-IHA = 5.7 vs. KPP90 = 2.9) and significantly greater in KPP120 than in KPP90 (KPP120 = 6.0 vs. KPP90 = 2.9). The SA/PM muscle activity ratio support our research hypothesis. Although there was no significant difference in the muscle activity of SA between KPPs, a significant difference in the muscle activity of PM seems to have affected the SA/PM muscle activity ratio. The ratio of muscle activity is a variable that can indicate the balance between interrelated muscles. The compensatory overactivity of the PM muscle due to a weakened SA muscle in individuals with scapular winging increases the compressive force on the glenoid [20] and narrows the subacromial space [21], which may result in shoulder joint instability or impingement pain. Therefore, in individuals with scapular winging, reducing the overuse of the PM muscle, as well as increasing the activity of the weakened SA muscle, should be considered an important factor. KPP120 in individuals with scapular winging increases the SA/PM muscle activity ratio more than KPP90 by decreasing the compensatory activity of the PM muscle. However, clinically, patients with scapular winging are often accompanied by limitations or pain in the shoulder joint at a shoulder flexion greater than 90° (e.g., adhesive capsulitis of the shoulder or impingement syndrome of the shoulder joint). Because these patients have difficulty performing KPP120, KPP90-IHA can be recommended to lower PM muscle activity and increase the SA/PM muscle activity ratio.

The SA/UT muscle activity ratio was significantly lower in KPP with IHA than without IHA, but there was no significant difference between shoulder flexion angles; this result does not support our research hypothesis. In this study, the muscle activity of UT during KPP was significantly greater at 120° of shoulder flexion than at 90° of shoulder flexion, but there was no significant difference in the SA/UT ratio according to the shoulder angle. SA, as well as UT, play an important role in the upward rotation of the scapula during KPP at 120° of shoulder flexion [10]. Although not statistically significant, the muscle activity of SA was increased in KPP120 when compared to KPP90. It can be interpreted that there was no significant difference in the SA/UT ratio according to different shoulder flexion angles because the SA muscle activity also increased as much as the UT muscle activity during 120° of shoulder flexion. Application of IHA in KPP90 can significantly decrease PM activity and significantly increase SA/PM ratio, but application of IHA in 90° of shoulder flexion increases unnecessary UT muscle activity and reduces the SA/UT ratio. Therefore, the application of IHA should be applied carefully during KPP where an individual has overused UT. The KPP120 also decreased PM and increased SA/PM ratio as compared to the KPP90, it was similar in results to the KPP90-IHA, but did not increase SA/UT. Therefore, we recommend that KPP120 be applied before applying IHA to KPP90 for individuals with scapular winging who have overuse of UT.

Our study has several limitations. First, it was difficult to generalize the results of our study to other groups because the participants included through the recruitment process comprised young men. The second limitation is that we did not measure the data of scapula kinematics experimentally. It is recommended that future studies determine whether scapular kinematic data can be used to increase SA muscle activity and decrease the compensatory overuse of the PM muscle. Third, since this study was cross-sectional, the long-term effects of the four different KPPs on the activity of the SA, PM, and UT muscles could not be determined. Thus, further studies should be conducted to evaluate the long-term effects of changes in the activity of the SA and PM muscles after KPPs. Lastly, since overuse of the PM in patient with scapular winging can be a major cause of shoulder joint disorders such as subacromial impingement or rotator cuff tear [2-4], so we recommended that future studies be conducted on patients with shoulder joint disorders with scapular winging.

We compared the effects of IHA on the activity of the SA, PM, and UT muscles and SA/PM and SA/UT muscle activity ratios during KPP at different shoulder flexion angles. When IHA was applied in KPP90, PM activity and the SA/UT muscle activity ratio decreased, and SA/PM muscle activity ratio increased. In KPP without IHA application, 120° of shoulder flexion yielded lower PM activity and higher SA/PM ratio than 90° of shoulder flexion. However, there was no additional effect in PM activity and SA/PM muscle activity ratio when IHA was applied at 120° of shoulder flexion. KPP90-IHA and KPP120 are the most effective exercises for decreasing PM activity and increasing the SA/PM muscle activity ratio in individuals with scapular winging. However, the application of IHA at 90° of shoulder flexion also decreases the SA/UT muscle activity ratio; therefore, it would be preferable to apply KPP120 to patients with the overuse of the UT muscle.

None.

None to declare.

No potential conflict of interest relevant to this study was reported.

Conceptualization: JHC, HSC, SMB, SHK. Data curation: JHC, SHK. Formal analysis: JHC, HSC, SMB. Investigation: JHC, HSC, SMB, SHK. Methodology: JHC, HSC, SMB, SHK. Project administration: JHC, HSC, SMB, SHK. Resources: JHC, HSC, SMB, SHK. Supervision: JHC, HSC, SMB, SHK. Validation: JHC, HSC, SMB. Visualization: JHC, HSC, SMB, SHK. Writing - original drafting: JHC. Writing - review and editing: JHC, HSC, SMB, SHK.