Delayed and disorganized muscle activation as a potential injury mechanism in men’s basketball jump shots

2.1 Participants

Ten male collegiate basketball players (age: 22.3 ± 1.25 years; height: 184.6 ± 3.06 cm; body mass: 75.1 ± 4.9 kg; training experience: 6.5 ± 1.6 years) and ten recreational basketball players (age: 21.9 ± 1.64 years; height: 181.9 ± 3.88 cm; body mass: 71.7 ± 5.0 kg) were recruited between 15 July 2024, and 30 May 2025 from a local university. The recreational players engaged in basketball activities two to three times per week and had no history of lower-limb or other musculoskeletal injuries prior to the study. Based on a G*Power 3.1.9.2 software (Heinrich Heine University Düsseldorf, Düsseldorf, NRW, Germany) analysis (α = 0.05, β = 0.80), the minimum detectable effect size was estimated to be f ≈ 0.33. G*Power analysis indicated that a total sample size of at least 20 participants was required to achieve adequate statistical power. The experimental protocol was approved by the Institutional Review Board of Jeonbuk National University, Jeonju, Korea (approval code: JBNU2024-09-015-002). Prior to participation, all participants were informed of the study procedures and provided written informed consent.

2.2 Procedure

During the experiment, each participant performed jump shots at two distances (5 m and 6.8 m). To ensure that both feet were in contact with the force plates at take-off, a movable basketball hoop was used, and the distance between the participant and the hoop was determined by adjusting the position of the hoop (Fig. 1). Prior to data collection, participants completed a standardized warm-up, including shooting practice, to familiarize themselves with the experimental setup. Based on biomechanical event markers, the shooting phase was defined according to ball position and body kinematic data and was divided into four stages: sinking stage, lifting ball stage, take-off stage, and release stage (Fig. 2) [21]. Each participant completed three valid trials at each shooting distance. For a trial to be considered valid, the jump shot had to be performed as a dynamic two-foot take-off jump shot; standing shots or shots executed without a jump after a run-up stop were excluded. Ball release was required to occur during the ascending phase of the jump or near the peak of the jump, and shots released during the descending phase were excluded. No additional adjustment steps, trunk rotations, or other movements that could alter shooting rhythm were permitted prior to the jump shot. A successful trial was defined as a smooth and continuous jump-shot motion without interruptions and without obvious deviations from typical competitive shooting mechanics. A rest interval of 2 minutes was provided between trials to minimize the effects of fatigue.

Fig. 1.Experimental setup. Jump shot distances are at 5 meters (middle) and 6.8 meters (long).

Fig. 2.Delineation of shooting phases.

2.3 Data collection

Kinematic data were captured using a 13-camera motion capture system (Motive 2.2.0, OptiTrack, Natural Point, Inc., Corvallis, OR, USA) operating at a sampling frequency of 240 Hz. Surface electromyography (sEMG) data were simultaneously recorded using a Trigno Wireless EMG system (Trigno Avanti, Delsys, Natick, MA, USA) with a sampling frequency of 1200 Hz. Reflective markers were placed on anatomical landmarks following the full-body Plug-in Gait model to construct a 7-segment biomechanical model in Visual3D (v2024.10.4, Charnwood Dynamics Ltd., Leicestershire, UK). The coordinate system was defined with the x-axis pointing forward, the y-axis pointing to the left side of the participant, and the z-axis oriented vertically upward.

Four right-side muscles were monitored: anterior deltoid (AD), biceps brachii (BB), triceps brachii (TB), and flexor carpi radialis (FCR). EMG electrodes were placed in accordance with the Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) guidelines. All EMG and motion data were synchronized via the acquisition system and stored for offline analysis.

2.4 Mathematical and statistical methods

A motion capture system was employed to process the video recordings of each participant’s jump shot trials. From these recordings, the most technically effective shot per athlete was selected for further analysis. Using the motion capture software, the three key phases of the shooting action—ball-holding and cushioning, push-off and take-off, and ball release—were identified and annotated by frame. The system subsequently exported the relevant motion data, providing the raw kinematic inputs necessary for biomechanical analysis of the shooting technique.

Surface EMG signals were processed to extract meaningful neuromuscular activation patterns. The raw EMG signals were first smoothed using a Savitzky-Golay filter, which attenuated high-frequency noise and enhanced the signal envelope for clearer temporal analysis. Following smoothing, all EMG data were time-normalized to 100 equidistant points, representing 1% to 100% of the total movement duration. This normalization ensured temporal consistency across trials and participants, facilitating reliable inter-group comparisons [31].

Muscle activation onset was detected using a windowed threshold method [32]. A muscle was considered active if its root mean square (RMS) amplitude exceeded 20% of the peak RMS value (within the selected trial) for at least five consecutive time points—corresponding to approximately 0.5 milliseconds in real-time. Although a minimum activation duration of 0.01 seconds was pre-defined, this was not applied under the time-normalized framework. To ensure temporal consistency of muscle activation and to reduce false-positive onset detection caused by signal noise or transient fluctuations, this study required the RMS amplitude to exceed the predefined threshold for at least five consecutive time points (≈0.5 s). Previous studies have demonstrated that incorporating a minimum duration criterion substantially improves the reliability and validity of EMG onset detection [33]. This thresholding approach, consistent with established Visual3D-based EMG processing protocols, ensured both reliability and reproducibility in identifying activation timing, thereby enabling robust cross-muscle and cross-group comparisons in neuromuscular control strategies. Each signal was processed using Savitzky-Golay filtering and subsequently normalized to the respective muscle’s peak RMS amplitude, facilitating inter-muscle temporal comparison. The X-axis represents a time-normalized shooting cycle divided into 100 evenly spaced intervals, corresponding to the full duration from preparation to follow-through.

For the assessment of SIS injury risk, this study primarily focused on temporal consistency of muscle activation. Temporal consistency was quantified using the standard deviation (SD) and the coefficient of variation (CV) of activation onset time (Eqns. 1,2):

A smaller SD indicates more stable and consistent activation timing, whereas a larger SD reflects greater unpredictability and variability in activation timing (Eqn. 3).

When the CV is less than 20%, temporal consistency is considered high; when the CV ranges from 20% to 40%, temporal consistency is considered moderate; and when the CV exceeds 40%, temporal consistency is considered low.

All statistical analyses were performed using IBM SPSS Statistics (v.27, International Business Machines Corp., Armonk, NY, USA). Initially, the sEMG RMS data were assumed to follow a normal distribution. The Kolmogorov-Smirnov test was applied to examine the normality of sEMG activation data under the 5 m and 6.8 m shooting conditions. The results indicated that all p-values were greater than 0.05, suggesting that the data met the assumption of normality and that the null hypothesis was accepted. Accordingly, independent-samples t-tests were used to compare between-group differences in sEMG data between high- and low-level players. Under the assumptions of normality and independence between groups, this parametric test was employed to analyze intergroup differences in sEMG RMS values of each muscle at different shooting distances.

3.1 5 m distance shooting

During the 5 m jump-shot task, the anterior deltoid (AD) of both high- and low-level players was activated during the lifting ball stage, with activation onset occurring at approximately 39%–45% of the normalized movement cycle (Fig. 3). The activation onset time of high-level players was later than that of low-level players. In contrast, the peak activation time of high-level players occurred earlier than that of low-level players, indicating a more advanced and temporally concentrated shoulder power output during the ball release phase.

Fig. 3.Anterior deltoid activation (5 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

Low-level players exhibited biceps brachii (BB) activation at the initial stage of the movement, with relatively pronounced activation during the lifting ball stage (Fig. 4). This activation subsequently decreased and then increased markedly again during the release phase. In comparison, high-level players primarily activated the BB during the lifting ball stage, with relatively stable activation during the ball-lifting phase and a significant increase only during the release phase.

Fig. 4.Biceps brachii activation (5 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

During the jump-shot task, no obvious differences were observed in the overall activation patterns of the flexor carpi radialis (FCR) between high- and low-level players. However, the activation onset time of high-level players (High Start) was significantly earlier than that of low-level players (Fig. 5).

Fig. 5.Flexor carpi radialis activation (5 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

The triceps brachii (TB) also exhibited differences between high- and low-level players during the 5 m jump shot. The activation onset time of high-level players (High Start) was earlier than that of low-level players (Low Start) (Fig. 6). The activation intensity of high-level players increased rapidly during the release phase and reached a peak of approximately 0.27% max, which was notably higher than that of low-level players (approximately 0.23% max).

Fig. 6.Triceps brachii activation (5 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

3.2 6.8 m distance shooting

During the 6.8 m jump-shot task, the AD of both high- and low-level players began to be activated in the middle to late stages of the movement. Compared with the 5 m condition, the overall activation of the AD shifted toward the later phase of the movement. The activation onset time of high-level players was earlier than that of low-level players (Fig. 7).

Fig. 7.Anterior deltoid activation (6.8 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

In the 6.8 m jump shot, low-level players exhibited BB activation at the initial stage of the movement and maintained a relatively high activation level during the lifting ball stage, followed by a slight decrease in the middle phase and a marked increase again during the release phase (Fig. 8). In contrast, high-level players showed low and stable BB activation during the first half of the movement, with pronounced activation mainly concentrated during the release phase. Moreover, the activation onset time of high-level players was clearly later than that of low-level players.

Fig. 8.Biceps brachii activation (6.8 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

During the 6.8 m jump-shot task, both high- and low-level players maintained low and stable activation levels of the FCR during the early to middle stages of the movement, with primary activation concentrated during the release phase. The activation onset time of high-level players was slightly earlier than that of low-level players (Fig. 9).

Fig. 9.Flexor carpi radialis activation (6.8 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

The TB demonstrated clear intergroup temporal differences during the 6.8 m jump shot. The activation onset time of high-level players (High Start) was earlier than that of low-level players (Low Start), and high-level players exhibited a steeper increase in activation during the release phase (Fig. 10).

Fig. 10.Triceps brachii activation (6.8 m shooting). Low Start: The position where activation begins at a low level. High Start: The position where activation begins at a high level. RMS: root mean square.

Based on the statistical results in Table 1, regardless of whether the jump shots were performed at 5 m or 6.8 m, no significant differences were observed between high- and low-level players in the activation sequence of the anterior deltoid, flexor carpi radialis, and triceps brachii. Only the biceps brachii showed significant differences between groups at both shooting distances.

3.3 Assessment of the risk of subacromial impingement (SIS) injury

Tables 2 and 3 presents the temporal characteristics and variability of surface EMG activation in four muscles during the 5 m jump-shot task for high- and low-level players. Under the 5 m shooting condition, clear differences were observed between groups in muscle activation timing variables associated with SIS.

Low-level players exhibited greater dispersion in the activation onset times of the AD, BB, and TB. Specifically, the standard deviations and coefficients of variation of activation onset times for these muscles were higher in low-level players than in high-level players, indicating lower temporal consistency of muscle activation.

Tables 4 and 5 presents the temporal characteristics and variability of surface EMG activation in the four muscles during the 6.8 m jump-shot task for high- and low-level players. Under the 6.8 m shooting condition, temporal differences in shoulder muscle activation became more pronounced. Low-level players exhibited greater variability in the activation onset times of the anterior deltoid, biceps brachii, and related synergistic muscles, with temporal consistency indices markedly lower than those observed in high-level players.

4.1 5 m distance shooting

Under the 5 m jump-shot condition, high- and low-skill players generally exhibited a similar sequential structure in the activation onset of the major upper-limb muscle groups; however, level-dependent differences were still evident in the timing of key muscles [37]. For the AD, both groups initiated activation during the ball-raising phase, indicating consistent temporal demands associated with shoulder flexion in short-distance jump shooting. Nevertheless, high-skill players demonstrated a slightly later AD onset but an earlier peak activation time compared with low-skill players, suggesting a tendency to concentrate shoulder output closer to the release phase without prematurely activating the shoulder musculature. This “later onset–earlier peak” temporal characteristic reflects a more streamlined and goal-directed neuromuscular control strategy.

The BB showed the most pronounced between-group temporal differences under the 5 m condition. Low-skill players exhibited early BB activation at the initial phase of the movement, maintained relatively high activity during the ball-raising phase, and demonstrated a second marked increase during the release phase, forming a premature and biphasic activation pattern, consistent with the findings of Wagner et al. [38]. In contrast, high-skill players primarily activated the BB during the ball-raising phase, with minimal early involvement, and only showed a pronounced increase near the release phase. This indicates that high-skill players avoid premature reliance on elbow flexors during short-distance shooting, thereby reducing unnecessary antagonist interference.

For the FCR, although the overall activation patterns were similar between groups, high-skill players exhibited a significantly earlier activation onset than low-skill players. This earlier recruitment of wrist stabilizers suggests that high-skill players begin preparing for fine wrist and hand control prior to release, whereas low-skill players tend to delay wrist activation until closer to the release phase.

The TB also demonstrated a trend toward earlier activation onset in high-skill players. As the primary driver of elbow extension, earlier TB activation facilitates smooth proximal-to-distal energy transfer during the release phase. In contrast, delayed TB onset in low-skill players may result in temporal compression of elbow extension during release. Kibler et al. [25] noted that early activation of wrist and elbow musculature provides a stable force-transmission platform for the release phase, rather than passive late-stage involvement.

Overall, during 5 m jump shooting, high-skill players exhibited later shoulder activation but earlier activation of distal stabilizing and propulsive muscles (FCR and TB), whereas low-skill players tended to activate the elbow flexor (BB) earlier, reflecting fundamental differences in movement organization strategies.

Although no statistically significant differences were observed in the overall muscle activation sequence between groups during 5 m jump shooting, analysis of activation timing and intermuscular coordination suggests that low-skill players may exhibit a higher potential risk for SIS.

Low-skill players demonstrated early BB activation during the initial phase of the movement and sustained elevated activity during the ball-raising phase. Previous studies have shown that the long head of the biceps tendon can exert a superior pull on the humeral head during shoulder flexion and elevation; persistent early activation may therefore increase superior humeral head translation and reduce the subacromial space [39, 40]. Such inappropriate temporal activation has been identified as an important neuromuscular risk factor for SIS.

In contrast, high-skill players delayed substantial BB activation until the ball-raising or release phase, while their AD activation, although initiated slightly later, peaked closer to the release phase. This strategy of minimizing premature antagonist involvement and concentrating shoulder output within critical phases contributes to maintaining shoulder joint dynamic stability [31]. Studies have shown that appropriate shoulder–scapular rhythm and muscle activation timing can effectively reduce mechanical compression of subacromial structures such as the supraspinatus tendon and subacromial bursa [41].

Furthermore, earlier activation of the FCR and TB in high-skill players indicates that distal stability and elbow extension drive are established prior to release, which may reduce compensatory loading at the shoulder joint during the terminal phase and indirectly lower subacromial impingement risk [42].

4.2 6.8 m distance shooting

Under the 6.8 m jump-shot condition, as shooting distance increased, the overall upper-limb muscle activation timing shifted toward later phases of the movement compared with the 5 m condition. This shift is consistent with the increased power output demands associated with long-distance shooting [43]. However, clear between-group differences in activation onset timing remained evident. The AD in both groups began activation during the mid-to-late phases of the movement, reflecting increased demands for late-phase shoulder force production during long-distance shots. Notably, high-skill players exhibited earlier AD activation than low-skill players, supporting the notion that long-distance shooting requires earlier organization of shoulder flexion output [44]. This suggests that high-skill players recruit shoulder flexors earlier to meet increased distance demands, whereas low-skill players demonstrate delayed shoulder involvement.

The BB remained the most discriminative muscle under the 6.8 m condition. Low-skill players again exhibited early BB activation during the initial movement phase, maintained high activity during ball raising, and showed further enhancement during release, indicating persistent and premature elbow flexor involvement. In contrast, high-skill players maintained low and stable BB activation during the first half of the movement, with marked activation only during release, and significantly later onset compared with low-skill players. Hirashima et al. [45] reported that premature elbow flexor activation disrupts proximal-to-distal sequencing and increases internal energy dissipation. These findings suggest that even under long-distance conditions, high-skill players suppress early elbow flexor involvement to avoid unnecessary interference within the shooting kinetic chain.

For the FCR, both groups maintained low activation levels during the early-to-mid phases, with substantial activation primarily occurring during release. However, high-skill players still demonstrated slightly earlier activation onset, indicating earlier preparation for wrist stabilization and force transmission during long-distance shooting.

TB activation timing exhibited more pronounced between-group differences under the 6.8 m condition. High-skill players not only initiated TB activation earlier, but also displayed a steeper increase during the release phase, indicating earlier and more concentrated elbow extension drive. Low-skill players showed delayed TB activation, potentially requiring elbow extension output within a shorter time window. This observation aligns with findings from baseball and basketball studies by Fleisig et al. [46] and Escamilla et al. [16], which suggest that skilled athletes initiate elbow extensor activation earlier to ensure temporal continuity and stability during release.

Overall, during 6.8 m jump shooting, high-skill players exhibited earlier involvement of shoulder and elbow extensor musculature combined with delayed elbow flexor activation, whereas low-skill players demonstrated premature elbow flexor involvement and relatively delayed proximal drive.

With increasing shooting distance under the 6.8 m condition, the demand for force output significantly increased, requiring the shoulder joint to perform under larger ranges of motion and higher loads, thereby amplifying the potential risk of SIS.

Low-skill players continued to exhibit early and sustained BB activation during 6.8 m jump shooting. This premature and persistent BB activity may further exacerbate superior shear forces on the humeral head during long-distance shooting, particularly under combined shoulder flexion and internal rotation, leading to significant compression of the subacromial space [47]. Reinold et al. [29] emphasized that poor neuromuscular control of elbow flexors and shoulder flexors during repetitive overhead activities is a key risk factor for subacromial impingement and rotator cuff pathology.

In contrast, although high-skill players initiated AD activation earlier to meet increased distance demands, they maintained delayed BB activation primarily concentrated during the release phase. This temporal adjustment indicates that high-skill players preserve a more appropriate division of labor among shoulder musculature under increased load, avoiding compensatory force generation via elbow flexors or anterior shoulder muscles [48].

Moreover, earlier initiation and rapid amplification of TB activation in high-skill players suggest that more kinetic energy is transferred through elbow extension and the distal kinetic chain, thereby reducing peak shoulder loading at release. Previous basketball and baseball studies have consistently shown that excessive concentration of power output at the shoulder joint without effective proximal-to-distal sequencing significantly increases stress exposure of subacromial structures [16].

Accordingly, during 6.8 m jump shooting, low-skill players may exhibit a higher potential risk of SIS due to premature BB activation and increased compensatory shoulder loading.

This study has several limitations. First, the sample size was relatively small and limited to male participants, which may restrict the generalizability of the findings. Second, the experimental task involved open jump shots without defensive pressure, which may not fully represent real-game conditions. Third, only four upper-limb muscles were analyzed, and the absence of trunk and lower-limb measurements limits interpretation of the kinetic chain.

Additionally, the use of a single “best trial”, rather than averaging multiple trials, may reduce reliability. Furthermore, the study primarily focused on activation onset timing without examining inter-joint coordination patterns.