Thirty-four subjects gave informed consent to participate. The study protocol was approved by Institutional Review Boards. Subjects had generally intact dentitions, and did not report or exhibit postcranial DJD, orofacial pain, gross asymmetries in craniomandibular anatomy as determined by examination, and were not pregnant as determined by medical history. Diagnostic classification was established by a clinical examiner using research diagnostic criteria for temporomandibular disorders28 (link) and a radiologist using magnetic resonance imaging and three-dimensional (3D) computed tomography.29 (link) The subjects, 18 females and 16 males, were divided evenly into two diagnostic groups (Table 1). Mean ages (SD) were 35 (14) and 34 (15) years for disc displacement and normal disc position groups, respectively.
A geometry file was created for each subject that described positions of the mandibular condyles, teeth, and five pairs of masticatory muscles (masseter, anterior temporalis, medial pterygoid, lateral pterygoid, anterior digastric), determined from standardized lateral and pos-teroanterior cephalographs according to a 3D coordinate system25 (link),27 (link) (Figure 1). Geometry files were used in a previously described numerical model,30 (link) first to validate the accuracy of the model in predicting data in each subject, and then to investigate inter-group differences in magnitudes of TMJ loads. Model-predicted ipsi-lateral and contralateral TMJ loads for a given static mandibular loading situation were resultant vectors at the anterosuperior-most mediolateral midpoint on the corresponding condyle and characterized in terms of magnitude and 3D orientation.
Model validation was determined by the ability to predict right and/or left sagittal plane projections of the TMJ stress-field trajectory in each subject31 (link) during symmetrical protrusion and retrusion of the mandible. That is, model-predicted orientations of TMJ loads were used as described previously and compared to individual-specific jaw tracking data measured in vivo.25 (link),27 (link),32 (link),33 Accuracy between model-predicted and measured data was deemed to be acceptable based on average errors of 16% (Iwasaki et al., personal communication). Then the validated model was used to predict magnitudes of TMJ forces per unit biting force (BF) using an objective function of minimization of muscle effort (MME).26 (link),34 The MME model calculated joint forces for biting on incisor, canine, and molar teeth, at a variety of angles (Tables 2, 3). Data were pooled and averaged by group. Analysis of variance (ANOVA) was used to test for significant differences between groups for magnitudes of TMJ loads during biting on incisors, canines, and molars at 13 angles.