An integrated multiscale analysis of injury mechanisms in sport impacts: an application to cervical spine biomechanics in rugby union scrummaging

Introduction and Objectives: By generating repetitive high-energy impacts (during the engagement phase) and intense
sustained loads (during the sustained push phase) under unstable conditions [1], the Rugby Union scrum has been
indicated as a possible risk factor for degenerative spinal injuries for forward players, and has been associated with
~40% of all catastrophic injuries in rugby [2]. However, little is known about how these external loading conditions
translate into internal stresses on the spinal structures. The aim of this study was to investigate the biomechanics of
cervical spine injury during rugby activities using an integrative approach: in-vivo and in-vitro experiments combined with
musculoskeletal modelling.
Methods: Three levels of analysis (Levels 1-3) were integrated. Level 1 was a biomechanical study of scrummaging (N=9
experienced rugby forwards) including motion capture (Oqus, Qualisys, Sweden), force measurement (force plates:
Kistler 92876BA, Switzerland; and instrumented scrum machine, [1]), and EMG of neck and trunk muscles (Delsys
Trigno, Delsys Inc, USA), carried out to assess the external kinematic and kinetic conditions acting on front row players.
The subsequent phases of scrummaging, initial engagament (impact and shock absorption) and sustained push, were
observed. Level 2 developed a bespoke musculoskeletal model (Rugby Model, [3]), consisting of 27 anatomical
segments, 26 joints, 78 cervical muscles, and 11 torque actuators, in OpenSim (OpenSim 3.2, SimTK, USA). The Rugby
Model was driven by the experimental data from Level 1 and was used to estimate joint dynamics, with a specific interest
in cervical joint motions and moments. Level 3 performed an in-vitro laboratory experiment to study the injury
mechanisms of porcine cervical spines subjected to impact loading conditions similar to those during scrummaging. Load
(2 load cells: SLC41/005000, RDP Electronics Ltd, UK) and deformations caused by impacts (mass of 12.86 kg dropped
from a height of 250 mm to give an impact velocity of ~2.2 m/s) were measured in a custom made impact rig, and highspeed
videos (2 Fastcam SA3, Photron Europe Ltd, UK) were used to investigate the mechanisms of injury through
digital image correlation (Vic-3D 2009.1.0, Correlated Solutions Inc, USA).
Results: Results from the biomechanical analysis confirmed that the load acting on the players, especially during the
initial engagement, was of a considerable magnitude (~2.8 kN compression force in single-player machine
scrummaging). Muscle activation patterns were affected by scrummaging conditions (e.g. machine vs. contested
scrummaging; ‘Crouch-touch-set’ vs. ‘Crouch-bind-set’ sequence) and phases of the scrum (e.g. pre-engagement vs.
engagement vs. sustaned push). For example, the activity of the erector spinae was significantly lower (in excess of 65%)
in machine scrummaging than in contested scrummaging, and the activation of sternocleidomastoid and upper trapezius
through pre-engagement and engagement were higher in the current ‘Crouch-bind-set’ technique than in the past ‘Crouchtouch-
set’ one. The computational musculoskeletal model highlighted an antiphase change in movement and loading
patterns between the upper and lower cervical levels (i.e. flexion load on the lower vertebrae and extension on the upper
vertebrae), and resulted in a “flattening” of the lordotic cervical curve during the impact phase. The present findings do
not provide direct evidence for injury mechanisms but seem in line with the patterns of injury that previous authors have
described in relation with scrum-related neck traumas [2]. The patterns of strain, load and resulting damages on the
cervical structures of the impacted porcine specimens were also similar to those clinically observed in injured players,
with the caudal vertebrae (C4-C6) more prone to damages (6 out of 8 specimens) as a consequence of the impact.
Fractures resulted from tension in the vertebral bodies due to first order buckling of the cervical spine in extension. The
mean maximum load in the cranial and caudal load cells was 5.8±2.0 kN and 6.0±2.1 kN and was reached at a time of
5.1±1.0 ms and 5.6±1.1 ms after impact, respectively.