Compensatory muscle activation and spinal curve changes in response to fatigue among adolescent male athletes

The optimal activity of several trunk muscles is required to adjust the posture and movement of the body even during different sports activities [1]. Although trunk function is only moderately related to strength and performance [2], it has the potential to predict the development of lower extremity overuse injuries [3]. It is therefore essential to assess trunk function as a part of an athlete’s fitness level; however, appropriate methods for assessing trunk function are not consistent in the literature [4‐7]. Most of these tests, including the prone plank (or prone bridge) test [7‐9], measure the holding time of a specific posture that has been often used to assess the strength and endurance of trunk muscles. The prone plank test is easy to repeat and requires no tools. However, evaluation is usually done only visually, in a subjective manner by a physiotherapist or other specialist. For example, the test personnel stops the measurement when the hip is perceived to move up or down five centimeters [10], or there are repeated, large deviations from the correct position [11]. Despite the test’s simplicity, results have shown a significant correlation with several athletic performance tests [10]. Adolescent children are going through a critical developmental period characterized by rapid biological changes that also affect core stability. However, considering that both postural control [12] and sagittal spinal-pelvic inclination [13] are superior in early adolescent girls than in boys, probably due to earlier maturation of muscle coordination [14], it seems that within a study only one gender should be focused on.

Surface electromyography (EMG) is often used to assess information about the neural activation of the muscles [15]. Previous studies aimed to determine the activation patterns of the corresponding muscles during a prone plank test via EMG [16‐18]. For example, van den Tillaar & Saeterbakken found similar rectus abdominis and obliquus externus abdominis muscle activation during a plank test until failure and 6RM back squats [18], supporting the idea that the prone plank test is an effective method to measure strength and endurance of trunk muscles. Although EMG data can serve as an objective tool for determining fatigue effects on muscle activation during the plank test, measuring the changes in kinematics is also essential to test trunk function [19].

Researchers often apply reflective markers on the skin to determine the position and shape of the spine in natural or specific settings for static or continuous movement [20‐23]. Briefly, the participant has reflective markers attached to the skin located at anatomical landmarks [24], i.e., over the spinous processes of different vertebrae, and the positions of the markers are determined by a motion capture (MoCap) system. Although skin marker measurements suffer from uncertainty, they can assess the changes in the lumbar and thoracic curvature angles reliably [23, 25]. A method to determine sagittal spinal curvatures based on skin markers has also been validated recently for static positions [26, 27], considering thoracic kyphosis (TK) and lumbar lordosis (LL). Since MoCap systems allow continuous recordings, the curvature values can be obtained for each timeframe.

Currently, there is no consensus on how to evaluate the prone plank test. We hypothesize that tracking the spinal curvature changes would offer an objective method to determine the outcome of the prone plank test while obtaining muscle fatigue data would help identify posture-related muscles that need further conditioning. Our goal is to develop and test an experimental protocol that has the potential to objectively evaluate the changes in spinal curves and muscle activation in response to fatigue during the plank test. To determine the degree of muscle fatigue, we calculated the decrease in median frequency (MDF) [28, 29].

Figure 3 shows the changes in spinal curves and the evolution of curvature values through time during the prone plank test from a representative participant. For the group, Fig. 5 shows the changes in TK and LL values. TK significantly increased from the first to the last 10 s of the plank test (T = 0; Z = -2.934; p = 0.003; r [Wilcoxon effect size] = 0.88), while LL angles either slightly reduced or greatly increased, which prevented us from detecting significant changes (T = 24; Z = -0.800; p = 0.424; r [Wilcoxon effect size] = 0.24) (Fig. 5, b). Although changes in TK and LL failed to show significant correlation (r_S = 0.573; p = 0.066), the scatterplot (Fig. 5b) clearly illustrates that they were strongly associated with each other besides the two outliers. However, there were no apparent connections between the initial values and the changes in response to the test (Fig. 5a).

The MDF of most of the measured muscles’ EMG activity showed various behaviors across the group (see example on Fig. 6); fluctuation around a constant value, steady decrease, and steady increase could all be observed during the prone plank test. Some muscles also exhibited a widely fluctuating MDF. In the participant group, these were typically the gluteus maximus, gluteus medius, gastrocnemius medialis, and biceps femoris. As an example, note the gluteus maximus, the gastrocnemius medialis, and the biceps femoris MDF in Fig. 6. The stationarity test indicated that 0.38% of epochs were non-stationary on average; the muscle with the highest level of non-stationarity was the erector spinae longissimus (2.67% of epochs on average).

A significant level of fatigue in response to the 60 s prone plank test (Table 1) was observed only in the case of the rectus abdominis (p < 0.001); the slope of fatigue was negative for all participants (Fig. 7). In the case of obliquus externus abdominis, nine of the eleven participants showed a negative MDF slope; however, the two positive values prevented us from finding a statistically significant non-zero fatigue (p = 0.102). The MDF slope values showed a wider range for the other measured muscles, encompassing both positive and negative values (Fig. 7).

Few significant correlations were found between muscle fatigue and spinal curve changes (Table 1). The level of fatigue for the rectus abdominis did not correlate with changes in either of the spinal curves (TK: r = 0.34, p = 0.30; LL: r = 0.01, p = 0.96), despite being the only muscle showing significant fatigue across the group. The increase in spinal curves significantly correlated with the fatigue of biceps femoris (TK: r = -0.75, p = 0.012; LL: r = -0.71, p = 0.019) (Table 1), indicating a compensatory muscle activation and spinal curve changes in response to fatigue.

The present study aimed to develop and test an experimental protocol that has the potential to objectively evaluate the changes in sagittal spinal and muscle activation in response to fatigue during the plank test using a MoCap system and EMG, respectively. Our results indicated that our recently validated method [26, 27] could be effectively used for tracking spinal curvature changes. A significant increase in TK angles from the first to the last 10 s of the plank test was found. We also determined the degree of muscle fatigue by calculating the MDF of the measured muscles’ EMG activity. Significant fatigue in response to the 60 s prone plank test was observed only in the case of the rectus abdominis; the fatigue of the other muscles varied considerably. Finally, the data indicated compensatory muscle activation and spinal curve changes in response to fatigue.

While TK significantly increased from the first to the last 10 s of the plank test in each participant during the 60 s prone plank test, changes in LL angles showed inconsistent results. Some of the participants straightened the lumbar area of the spine which corresponded to slightly reduced LL angle values, while others greatly increased LL angles by tilting the pelvis, suggesting that participants used different strategies in response to the demands of the test. This is in line with a previous study [11] that observed several manifestations of fatigue during a plank test, including shaking, back pain, and abdominal and thigh fatigue. The observed great increase in LL may be an indicator of failing to complete the prone plank test in case of fixed time duration or may mark the end time of successful execution. Consequently, the application of a MoCap system during the prone plank test can be a useful tool to objectively determine the outcome of the postural muscle test based on spinal curvatures and the motor control strategies in response to fatigue in order to detect trunk function.

In the present study, MDF for most muscles showed various behaviors (increase, decrease, or constant with slight fluctuations) across the group. In addition, gluteus maximus, gluteus medius, gastrocnemius medialis, and biceps femoris tended to show largely fluctuating activities during the test. The statistical analyses on the MDF slopes indicated a significant level of fatigue for the rectus abdominis muscle (p < 0.001) in response to the 60 s prone plank test, while the same for the externus obliquus abdominis was not significant (p = 0.102). This is not in line with the study from Czaprowski et al. [34], which found that the obliquus externus abdominis muscle had the highest EMG activity (42.3 ± 19.5% of MVC) during a prone plank test. Nevertheless, Escamilla et al. [35] showed an almost identical muscle activation for the rectus abdominis (40 ± 10% of MVC) and obliquus externus abdominis (40 ± 20% of MVC) during a plank test, suggesting that both muscles play a significant role in maintaining the plank position and are essential for trunk function. Further studies should determine whether their activation level was associated with sports performance.

Although only the changes in the MDF of rectus femoris correlated with the changes in spinal curves, our results support the idea that the proposed experimental setup has the potential to objectively determine the fatigue-induced changes in the kinematics and muscle activation during a prone plank test using a MoCap system and EMG, respectively. For example, a substantial change in the TK curve at ~ 40 s in Fig. 3 can be explained by the decreased activation of gluteus maximus, biceps femoris, and erector spinae iliocostalis (Fig. 6). It is possible that during a 60-second-long exercise, the activation of the involved muscles within a muscle group (e.g., abdominal muscles, hip flexors, hip adductors, knee flexors, etc.) changes through time to reduce fatigue. Therefore, grouping the muscles by function and analyzing their activity together are recommended to clearly detect whether changes in muscle activation in response to fatigue are associated with spinal curve changes.

Limitations arise from the small sample size and the use of surface electrodes which prevented the detection of the activity of deeper posture-related muscles. Future studies need to recruit more participants of different ages and sports. Another limitation was the test’s fixed duration, which may not have induced fatigue in some individuals. This can be avoided by testing until failure. Our experimental setup could also be applied in other trunk function tests, e.g., the Matthias test, and can involve other joints. Moreover, grouping the muscles by function and analyzing their activity altogether may allow us to detect the participants’ voluntary and involuntary motor control strategies. Future studies may also consider determining the role of shoulder and arm muscles when increased kyphosis due to fatigue is present during the prone plank test.

In conclusion, we found a significant increase in thoracic kyphosis angles from the first to the last 10 s of the plank test. Lumbar lordosis angles slightly decreased for some participants while markedly increased for others. Consistent and significant fatigue was observed across the group only in the case of the rectus abdominis, but the level of fatigue did not correlate with spinal curve changes. However, biceps femoris fatigue significantly correlated with the changes in kyphosis and lordosis angles to a large degree. Moreover, the data indicated compensatory muscle activation and spinal curve changes in response to fatigue. These preliminary results suggest that the proposed method has the potential to objectively detect the strength of posture-related muscles during the prone plank test. Our experimental protocol may support future researches that aim to objectively detect the strength of posture-related muscles via any other trunk function test at any age group and sport and identify muscles that need further conditioning on an individual level.