Validation of diagnostic criteria for sleep bruxism

Abstract

Several diagnostic criteria for bruxism can be taken from the literature; however, most of them have never been validated. This study examined whether predictor variables taken from physical examinations and questionnaires were related to the actual bruxism levels. Fifty dental students agreed to participate in this study and eight examination variables and seven questionnaire variables were collected from them. The subjects measured their nocturnal EMG activity from the right masseter muscle for six consecutive nights in their home by means of a portable EMG device. Off-line analysis was performed on data from second to sixth nights. By using a custom made software, all EMG activity elevations above a minimum threshold of 50% of each subject's individually established maximum voluntary contraction (MVC) level were quantified with regard to the duration and number of elevations and then three outcome variables, which were event number per hour (number/h), event duration per hour (duration h−1), and duration per event (duration/event), were calculated. A multiple stepwise regression (MSR) analysis was conducted to assess the 15 predictor variables and the three outcome variables. These MSR analyses revealed that the joint sound score remained in the regression equation as a predictor (n=50, P < 0·05) of the likelihood that a subject would exhibit longer bruxism events (duration h −1 and number h −1 ). It must be noted that the self-awareness and tooth attrition status were found not to be strong predictors and even for the above variable where significant association was found, the likelihood ratio between the variable and predicted outcomes was not robust.