Introduction:
The horse is a quadrupedal animal without a collarbone, which means that the only rigid connection between the appendicular skeleton and the axial skeleton is via the hindlimbs that articulate with the pelvis through the coxofemoral joint. While hindlimb motion in the horse is basically in the sagittal plane only, the motion of the equine spine is more complex and comprises flexion- extension (which can be technically described as a rotation around an axis perpendicular to the sagittal plane, also called 'pitch' in aeronautical terms), lateral flexion (rotation around the vertical axis or 'yaw'), and axial rotation (around the axis pointing in the direction of movement, or 'roll'). The 'merger' of these different motion patterns takes place in the sacroiliac joint, which is therefore a crucial structure and, possibly for this reason, often involved in clinical problems with 'sacroiliac disease' (SID) being a frequent (though not always well-defined) disorder in performance horses.
Due to its difficult accessibility and the low ranges of motion between different parts of the pelvis, studies on pelvic biomechanics have been very limited thus far and substantially less work has been done on pelvic biomechanics than on other parts of the equine axial skeleton. There is, however, growing insight that, as in human beings, pelvic biomechanics and disturbances therein are of great importance for athletic performance.

Functional anatomy:
The equine pelvis is connected to the appendicular skeleton by the 2 coxofemoral joints, which have a large range of motion and suffer surprisingly few clinical problems in the horse, certainly if compared with some other species (dogs, humans). It is connected with the axial skeleton through the lumbosacral junction, which can be either the articulation between the 6th lumbar vertebra and the first sacral vertebra, or between the 5th and 6th lumbar vertebrae if the last has fused with the sacrum (has 'sacralised'), as is frequently seen in various breeds of horses, including the Thoroughbred. The range of (flexion-extension) motion in the lumbosacral junction is relatively large (about 23degrees, Degueurce et al. 2006) and is much larger than of any other intervertebral joint, with the exception of the cervical vertebrae.
To understand the biomechanics of the equine pelvis and the biomechanical background of many forms of SID, an insight of motion within the pelvis itself is much more important than knowledge of motion type and range of the interfaces with the limbs and the thoracolumbar spine. As said, the sacroiliac joint is crucial to pelvic function. This joint has a very specific configuration, with one side of it (the sacral side) covered with hyaline cartilage, the other (iliac) side with fibrocartilage. This hybrid form between a diarthrodial joint and a fibrous connection is most probably the answer to the type of loading the joint is subjected to. There is hardly or no compressive loading (as in most diarthrodial joints), but the sacroiliac joint is mainly subjected to torsional and shear forces. This can also be deduced from the ranges of motion that have been measured in ex vivo experimental set-ups. Movement in the sagittal plane in the sacroiliac joint is designated as nutation (which is an increase of the distance between the ischial bone and the sacrum) and counternutation (a decrease in the same distance).

This movement is (other than the comparable flexion-extension movement in many intervertebral joints) extremely limited and has been measured at approximately one degree (Degueurce et al. 2006; Goff et al. 2006). However, lateral motion and torsion in the coronal and axial planes respectively was larger (approximately 2.5degrees, Goff et al. 2006). In fact, the movements are coupled and, as in the other parts of the axial skeleton, there is no lateral bending without axial rotation and vice versa. In a later study, Haussler et al. (2009) measured even larger values (up to approximately 8degrees) for angular rotation when applying a well-defined mechanical force (as opposed to manual force used in Goff's study).
Apart from the motion within the joint, it is also important to note that the pelvis itself, although being a bony structure, deforms as well. In an ex vivo study Haussler et al. (2009) showed that asymmetric pelvic deformation occurs during most sacroiliac joint movements.

Proprioceptive control of motion:
The pelvis consists of more than bones and articulations alone. There are various strong ligaments attached to the pelvic bones, some of which have a clear role in pelvic biomechanics. Degueurce et al. (2006) showed that after desmotomy of the broad sacrotuberous ligament or sacrosciatic ligament (running from the lateral aspect of the sacrum and the transverse processes of the first 2 caudal vertebrae to the ischiatic spine and tuber ischium) and the sacrotuberal or dorsal sacroiliac ligaments (2 portions, running from the sacral tubers to the sacral spinous processes and the lateral sacral crest respectively) nutation almost doubled to 1.7degrees. The ligaments may have more than a mechanical function only, as they are richly innervated by proprioceptive nerve endings. There is growing evidence in the human species that the force transfer in the region of the sacroiliac joint is under tight proprioceptive control of the neural elements in the ligamentous structures surrounding the joint, effectuated by especially the deep musculature (Goff et al. 2009). There is no reason to believe that this situation is different in the horse.

Conclusion:
Pelvic biomechanics can still be considered a weakly researched area in the horse. This is due to the difficult accessibility of the region and the small ranges of motion, which makes especially in vivo work difficult. Nevertheless, although motion in the sacroiliac joint may be limited, proper functioning of this joint is crucial to unhampered locomotion. Further research into the proprioceptive control and possible aberrations therein of pelvic biomechanics is warranted and may yield clinically very relevant information.