The goal of this study was to quantify the relative contributions of each muscle group surrounding the spine to vertebral joint rotational stiffness (VJRS) during the push-up exercise. Upper-body kinematics, three-dimensional hand forces and lumbar spine postures, and 14 channels (bilaterally from rectus abdominis, external oblique, internal oblique, latissimus dorsi, thoracic erector spinae, lumbar erector spinae, and multifidus) of trunk electromyographic (EMG) activity were collected from 11 males and used as inputs to a biomechanical model that determined the individual contributions of 10 muscle groups surrounding the lumbar spine to VJRS at five lumbar vertebral joints (L1-L2 to L5-S1). On average, the abdominal muscles contributed 64.32 ± 8.50%, 86.55 ± 1.13%, and 83.84 ± 1.95% to VJRS about the flexion/extension, lateral bend, and axial twist axes, respectively. Rectus abdominis contributed 43.16 ± 3.44% to VJRS about the flexion/extension axis at each lumbar joint, and external oblique and internal oblique, respectively contributed 52.61 ± 7.73% and 62.13 ± 8.71% to VJRS about the lateral bend and axial twist axes, respectively, at all lumbar joints with the exception of L5-S1. Owing to changes in moment arm length, the external oblique and internal oblique, respectively contributed 55.89% and 50.01% to VJRS about the axial twist and lateral bend axes at L5-S1. Transversus abdominis, multifidus, and the spine extensors contributed minimally to VJRS during the push-up exercise. The push-up challenges the abdominal musculature to maintain VJRS. The orientation of the abdominal muscles suggests that each muscle primarily controls the rotational stiffness about a single axis.
Samuel J. Howarth, Tyson A.C. Beach and Jack P. Callaghan
Stefan Sebastian Tomescu, Ryan Bakker, Tyson A.C. Beach and Naveen Chandrashekar
Estimation of muscle forces through musculoskeletal simulation is important in understanding human movement and injury. Unmatched filter frequencies used to low-pass filter marker and force platform data can create artifacts during inverse dynamics analysis, but their effects on muscle force calculations are unknown. The objective of this study was to determine the effects of filter cutoff frequency on simulation parameters and magnitudes of lower-extremity muscle and resultant joint contact forces during a high-impact maneuver. Eight participants performed a single-leg jump landing. Kinematics was captured with a 3D motion capture system, and ground reaction forces were recorded with a force platform. The marker and force platform data were filtered using 2 matched filter frequencies (10–10 Hz and 15–15 Hz) and 2 unmatched filter frequencies (10–50 Hz and 15–50 Hz). Musculoskeletal simulations using computed muscle control were performed in OpenSim. The results revealed significantly higher peak quadriceps (13%), hamstrings (48%), and gastrocnemius forces (69%) in the unmatched (10–50 Hz and 15–50 Hz) conditions than in the matched (10–10 Hz and 15–15 Hz) conditions (P < .05). Resultant joint contact forces and reserve (nonphysiologic) moments were similarly larger in the unmatched filter categories (P < .05). This study demonstrated that artifacts created from filtering with unmatched filter cutoffs result in altered muscle forces and dynamics that are not physiologic.
Lindsay L. Musalem, Tatjana Stankovic, Drazen Glisic, Gillian E. Cook and Tyson A.C. Beach
The objective of this study was to investigate why holding times on 2 different tests of isometric trunk flexor endurance capacity (prone plank and v-sit) are weakly correlated. Body position and ground reaction force data from 10 men and 10 women were used to conduct static biomechanical analyses of both test postures, and bilateral activations of the rectus abdominis, internal and external obliques, latissimus dorsi, and lumbar and thoracic erector spinae were measured in a second sample of 15 men and 15 women while holding the test postures. No between-posture differences in net low back flexor moments were found (P = .111), but the lumbar spine was 28° more flexed in the v-sit than in the plank (P < .001). No between-posture differences were detected in the rectus abdominis (P = .397), external obliques (P = .204), internal obliques (P = .226), or lumbar erector spinae (P = .116) activation levels, but those of the thoracic erector spinae (P = .0253) and latissimus dorsi (P < .001) were greater in the plank than in the v-sit. Altogether, the findings suggest that differences between plank and v-sit holding times are most likely related to between-test differences in lumbar spine postures and shoulder demands.