Multifactorial comparison of disk displacement with and without reductionto normals according to temporomandibular joint hard tissue anatomic relationships

Statement of problem. There is disagreement about the predictive value of temporomandibular joint tomographic anatomy in the diagnosis of internal derangements. Purpose. This study aimed to identify multifactorial temporomandibular hard tissue relationships that differentiate disk displacement with reduction and disk displacement without reduction from normals. Material and methods. Temporomandibular joint tomograms from females diagnosed with unilateral disk displacement with (n=84) or without (n=78) reduction were compared to 42 asymptomatic normal joints with the use of 14 linear and angular measurements and 8 ratios. A validated classification tree model was tested for accuracy with sensitivity, specificity, goodness of fit, and the amount of log likelihood accounted for. The tree model was compared with a multiple logistic regression model and univariate testing. Results. The disk displacement with reduction tree model consisted of 3 disease and 2 normal pathways with interactions between fossa width to depth ratio, condyle position, and linear posterior joint space. This class was characterized by either a much wider- and shallower-than-average fossa shape and/or by a moderately posterior condyle position when the fossa shape was average to deeper and/or narrower. The logistic regression and univariate models also suggested wider and/or shallower fossae, as well as longer eminence length. The disk displacement without reduction tree model consisted of 2 disease pathways and 1 normal pathway. Interactions characterized this class by either a posterior to very posterior condyle position or by a much deeper than average fossa depth when the condyle position was concentric to anterior. The logistic regression model emphasized greater fossa depth and width versus normals. The tree models conservatively predicted the disease classes: Rescaled Cox and Snell R² 37.0%, sensitivity 70.2%, and specificity 90.5% for disk displacement with reduction; R² 28.8%, sensitivity 66.7%, and specificity 85.7% for disk displacement without reduction. Conclusion. Within the limitations of this study, hard tissue relationships revealed by central tomogram sections were able to model notable differences between disk displacement with reduction and disk displacement without reduction versus asymptomatic normals when temporomandibular joints were examined as a multifactorial system typified by interactions of fossa width to depth proportions and condyle position. While substantial, the hard tissue predicted only part of the biology. The model could be broadened by additional factors and interactions.