Comparison of Elbow Joint Forces and Torques between Youth and Elite Pitchers: A Statistical Parametric Mapping Approach

Introduction: Research has shown similar motions and timing between youth and adult pitchers with differences noted in elbow kinetics. Reports show the rates of UCL injuries has been rising in recent years in both youth and adults. Inverse dynamic pitch models can estimate external elbow torque and joint forces to serve as proxies for UCL loads. While magnitude differences in elbow forces and torques between youth and adult pitchers have been reported, peak values only provide a discrete description of the loads. Moreover, there is evidence that the use of impulse values may not always reflect the true meaning of these loads in the time domain. Hence, the nature of where and how these loads differ between youth and adults over a time series domain may prove to be a more accountable analysis of elbow loads during the pitch.

Pitching 3D kinetic time series data at the elbow are complex. Statistical Parametric Mapping (SPM) performs topological inferences over time series data sets instead of performing inferential tests at select and discrete time points. Simply, SPM affords the ability to visualize the statistical results (a time series of t-values) in the time series domain of the pitch.

The purpose for this study was to evaluate the differences in elbow force and torque profiles over the pitch sequence between youth (under age 16) and adult (consisting of collegiate and professional) pitchers throwing a fastball.

Materials and

Methods: 14 youth (mean age 14.0±0.75 yrs; mean height 58.5±3.18 cm; mean weight 96.5±15.04 kg) and 11 collegiate athletes (mean age 22.3±3.24 yrs; mean height 75.2±2.12 cm; mean weight 188.5±18.5 kg) performed maximal velocity pitches for strikes from the stretch position. Full body, 3D marker coordinates were collected with a 16-camera mocap system (200 Hz). A radar gun recorded pitch velocity. The 3 highest velocity pitches for a strike were selected for analysis. Coordinate data were filtered with a low pass cutoff frequency of 16Hz, and 3D marker trajectories were input into a standard inverse dynamics routine that produced upper limb joint forces and torques according to ISB guidelines. The elbow joint force profiles for each pitcher were normalized to participant’s body weight (%BW) where; FX= med-lat, FY=ant-post, FZ=tension-compression. Elbow torque conventions yielded TX=ext-flex, TY=varus-valgus, TZ=sup-pronat. Joint torques were normalized to body weight and height (%BW*Ht). The time frame of the pitching motion was normalized and defined as ball separation from the glove (0%) to point of maximal internal rotation of the shoulder (MIR =100%). The pitching events were defined by stride foot contact (SFC-40% of pitch cycle), maximum external rotation (MER-80% of pitch cycle) and ball release (REL-90% of pitch cycle). REL was defined by observing trajectories of markers on ball as it leaves the hand (ball-hand marker separation in X-plane >20 mm). The time-normalized data series were then used as input for the SPM analysis implemented in MatLab using SPM1D, a free and opensource package (a priori, p< 0.05).

Results, Conclusions, and Discussions: Adult pitches threw significantly faster (37.3±1.9 m/s; 85.5 mph) compared to the youth (20.7±3.45 m/s; 47.3 mph). Figure 1 below identifies where within the pitch cycle differences occurred. The first 0- 40% of the pitch cycle is ‘clipped/omitted’ from the figure as no differences were noted in that time frame (all p≥.05). Differences in elbow flexion-extension (Elb Ang) noted youth maintained a significantly less flexed arm starting from 73% of the phase up through 82%. But both groups released the ball at nearly the same elbow angle position (both within 5 degrees of full extension). Most differences in elbow joint reaction forces (Elb FX, FY FZ) occur between MER and REL (between 80% - 90%). Notably, adult pitchers exhibited higher FX-medially directed forces (forces that would act on the UCL) starting earlier (at 68%) of the pitch cycle; but the absolute magnitude of the differences was not significant at that time point when adjusted to %BW. This suggests that the pitch performance exhibited by adults begins loading the UCL much earlier in the pitch and with ever increasing loads until FX peaks at MER at 40%BW (300-400N). Conversely, FY-anteriorly directed forces and FZ-compression forces did not differ until just at MER. of the pitch cycle. No differences were noted in any of the elbow torque values when torque values were adjusted to %BW*Ht.

In conclusion, SPM can be an important tool to visualize topological inferences over the pitch cycle and allows for easily identifiable time points that designate when and for over what duration these differences can occur. Moreover, these differences are not readily apparent when using discrete values abstracted from a time series set alone. Despite the large variability in pitchers, differences in force profiles occurred during relatively short periods of the entire pitch cycle between youth and adult pitchers and these occurred most notably between the MER and REL phases.

Acknowledgements (Optional):