This paper “ISB recommendations on the reporting of intersegmental forces and moments during human motion analysis” [] provides an excellent contribution to clarify several issues about musculoskeletal intersegmental and joint biomechanics. Notably, it addresses the often confused concepts of ‘internal’ and ‘external’ moments.
I do have some comments about the following passage in the manuscript, in relation to spinal biomechanics:
Joint centers
To compute the intersegmental joint moments, a reduction point, that is the point with
respect to which the system of forces is reduced, is required. This point is classically defined as a
joint center. In most of the human movement analysis protocols proposed in the literature,
adjacent bony segments are conceptually assumed to be connected by spherical pairs, and their
relative motion is described by three joint angles about the three anatomical axes defining the
joint coordinate system and passing through this joint center.

My own experience of joint biomechanics relates mostly to the spine, where the articulations are flexible structures, not simple diarthroidial joints, so they transmit moments in addition to forces. Therefore, especially for the spine, “Joint Center” would benefit from further definition. The reason it is “classically defined as a joint center…. [and] adjacent bony segments are conceptually assumed to be connected by spherical pairs” is that the joint contact force (in the case of a frictionless joint) passes through the center of both spherical surfaces, i.e. the joint force is perpendicular to the surface. (This is qualified by ‘in most of the human movement analysis protocols’.) Frequently, the joint center is identified as the (kinematic) center of rotation. This is convenient, since for a frictionless, rolling and gliding joint, the center of rotation is located on a line that is perpendicular to the surface at the point of contact, i.e. on a line co-linear with the joint force.

This concept is often (incorrectly) extended to the spine articulations. In the spinal case the instantaneous center (or axis) of rotation depends on the present or actual forces transmitted by the (flexible) articulation. There is no a priori reason why the articulation would not be transmitting moments about this instantaneous point or axis. Therefore, my colleague Mack Gardner-Morse and I have employed an ‘equivalent’ structure (at its simplest an ‘equivalent beam’) to represent the spinal articulations. The properties approximate those obtained from experimental studies. See Gardner-Morse MG, and Stokes IAF: Structural behavior of human lumbar spinal motion segments. J Biomechanics 2004, 37(2): 205-121