Linear Regression Analysis of Predictors of Temporomandibular Joint Disorders

Temporomandibular disorders (TMD) are defined as a group of musculoskeletal conditions that affect the masticatory system and are of multifactorial etiology cause. Many factors have been discussed in the literature as possible etiologies of TMD. The diagnosis of TMD requires accurate diagnostic predictors. Objective: The purpose of this study was to identify the subjective and objective predictors that may classify a subject as having TMD. Methods: Three hundred and forty-five subjects completed a subjective questionnaire and had a clinical and dental examination for signs and symptoms commonly associated with TMD and/or disc displacement (DD). The subjects were divided into two groups: eighty-two asymptomatic controls and 263 symptomatic TMD patients. The symptomatic subjects were further divided as having DD, unilateral or bilateral (N=221), or having bilateral symptomatic but normal joints (N=42). Sixty variables were subjected to linear regression analysis. Paired T-tests were performed to identify the predictive variables in asymptomatic controls and symptomatic TMD patients. The post-ANOVA analysis was performed comparing GROUP (asymptomatic controls vs. symptomatic subjects) vs. DIAGNOSIS (normal joint vs. DD) to identify the best predictors of TMD. When GROUP and DIAGNOSIS were significant for a variable; this was considered a good predictor of TMD. Results: The presence of joint sounds (clicking), right anterior Crossbite, positive muscle palpation (superficial masseter, middle temporal, posterior temporal and lateral TMJ), deviation of the mandible on opening, jaw joint pain, maxillary midline deviation, and limited left lateral movement were significant (p<0.05) for GROUP and DIAGNOSIS. Conclusion: This study suggests that of the sixty varibles studied, only 10 may be accurate predictors to classify a patient as having TMD.