Related Experiment Video
Updated: Jul 16, 2026
Isokinetic Robotic Device to Improve Test-Retest and Inter-Rater Reliability for Stretch Reflex Measurements in Stroke Patients with Spasticity
Published on: June 12, 2019
Examination of the Construct Validity of Nextiles Arm Sleeve as a Measure of Peak Elbow-Varus Torque
Jakob Wolf1, Kristen Nicholson1, Garrett Fernandez1
1Department of Orthopaedic Surgery & Rehabilitation, Wake Forest School of Medicine, Winston-Salem, NC.
Context:
New wearable circuited fabric biomechanic technology that can be used in a more ecological context to measure elbow-varus torque and throw count is now available but needs to be validated compared with markerless biomechanics.
Objective:
To examine the construct validity of the Nextiles Arm Sleeve by assessing its agreement and convergence with markerless 3-dimensional (3D) biomechanics used as a reference for (1) peak elbow-varus torque and (2) pitches thrown with separate analyses for fastballs, breaking balls, and changeups.
Design:
Cross-sectional study.
Setting:
Laboratory.
Patients Or Other Participants:
A total of 29 collegiate pitchers (age = 19.5 ± 1.3 years, height = 1.88 ± 0.06 m, mass = 91.7 ± 9.5 kg, body mass index = 26.2 ± 1.9 kg/m2) participated.
Interventions:
Pitchers were assessed simultaneously via the Nextiles sleeve and markerless 3D motion capture (KinaTrax) as they threw fastballs, breaking balls, and changeups to a catcher at regulation distance.
Main Outcome Measures:
Nested Bland-Altman limits of agreement (LoA) for construct validity, mixed-effects linear regressions for convergence validity, and intraclass correlation coefficients (ICCs) for test-retest reliability were calculated between the Nextiles sleeve and 3D biomechanics for elbow-varus torque. We determined that an a priori acceptable threshold for LoA and convergence was a 95% CI width of 0.15 and 95% CI lower limit of 0.50.
Results:
Pitchers threw 200 pitches (fastball: 72%, n = 143; breaking ball: 14%, n = 27; changeup: 15%, n = 30), with 182 pitches recorded by the Nextiles sleeve and 18 pitches (n = 3 pitchers) not recorded due to technical errors. Agreement and validity for pitch measures (mean difference [lower, upper] LoA = 6.7 N·m [-26.2, 39.6 N·m]; β = 0.31 [0.12, 0.52]; Nextiles sleeve ICC [95% CI] = 0.96 [0.93, 0.98]; KinaTrax ICC [95% CI] = 0.97 [0.94, 0.98]) were higher compared with agreement for body weight × height (BW×H; mean difference [lower, upper] LoA = 0.37 BW×H [-1.41, 2.45 BW×H]; β = 0.37 [-1.41, 2.15]; Nextiles sleeve ICC [95% CI] = 0.93 [0.88, 0.96]; KinaTrax ICC [95% CI] = 0.95 [0.92, 0.98]). We found no interaction per pitch type (P = .47) or pitch thrown (P = .58).
Conclusions:
The Nextiles sleeve collected pitch counts accurately except for 3 pitchers for whom it did not collect any pitches thrown due to technical errors. However, the demonstrated LoA and convergence were beyond the a priori acceptability threshold and lower than the a priori defined acceptable lower bound of the 95% CI. These results suggest that the Nextiles sleeve does not measure the same construct as KinaTrax.
