Association between shoulder proprioceptive control and pitching accuracy in high school baseball players: an exploratory position-stratified analysis

Pitching accuracy is critical for performance and injury prevention, yet its neuromuscular determinants remain poorly understood. This study explored the relationship between active shoulder proprioceptive precision, quantified via 3D Movement Variability (3DMV), and multidimensional pitching accuracy in high school baseball players. Eighty-six high school baseball players participated. Active joint repositioning tests quantified 3DMV of the scapulothoracic and glenohumeral joints across pitching-specific phases (cocking and ball release). Pitching accuracy was assessed using absolute error, constant error, and 95% confidence ellipse parameters (area, radii). Relationships were analyzed using Pearson’s correlations and Canonical Correlation Analysis (CCA) to identify multivariate associations. Scapular 3DMV significantly correlated with pitching accuracy during the cocking and ball-release phases ( r = -0.227 to -0.244, p < 0.05). Humeral 3DMV at ball release was associated with pitching dispersion ( r = 0.236, p = 0.03). While global CCA was not significant ( Rc 1 = 0.399, p = 0.40), subgroup loading patterns suggested different descriptive tendencies between playing positions. Pitchers showed a relatively higher loading on humeral control during the cocking phase (loading = 0.91), whereas fielders demonstrated higher loading on scapular stability (loading = 0.77). No significant group interactions were observed in the fundamental sensory–motor relationship. Exploratory position-stratified analyses suggest potential differences in shoulder proprioceptive control that may be related to task-specific demands in high school baseball players. Subgroup loading patterns suggested relatively higher humeral loadings during the cocking phase in pitchers and higher scapulothoracic stabilization-related loadings in fielders. However, these observations were not supported by significant interaction effects. Accordingly, the findings should be interpreted as exploratory and hypothesis-generating, suggesting potential variations in sensorimotor control characteristics across playing positions. These results may provide preliminary insights to guide future research on position-related training considerations, pending confirmation in adequately powered studies. Clinical trial registration: This study was registered at ClinicalTrials.gov Protocol Registration and Results System (NCT06704529) (https://clinicaltrials.gov/study/NCT06704529). The data presented here constitute a secondary analysis focusing on position-specific proprioceptive control.