Sensitivity and specificity in a longitudinal temporomandibular disease study

A longitudinal prospective study on the development of temporomandibular disease (TMD) in children was initiated in 1970. The study ended in 1990. One of several diagnostic TMD criteria applied was osteoarthrosis, defined as a 'morphological and/or structural deformity of the condyle' diagnosed on radiographs. Although these radiographic techniques are suspect in terms of falsenegative findings, 'morphological and/or structural deformity of the condyle' has been shown to be associated with a characteristic craniofacial form. As such, it has confirmed diagnostic significance. The present study will test the reliability of this diagnosis using cross-sectional and longitudinal findings for objective signs and subjective symptoms that are attributed to TMD. Signs and symptoms of TMD performed poorly as predictors for 'morphological and/or structural deformity of the condyle'. Thirty-seven percent of the participants had a 'morphological and/or structural deformity of the condyle' diagnosed at least once during the 20-year study period. In two thirds of the cases, 'morphological and/or structural deformity of the condyle' was not a stable characteristic through time. Normal radiographic anatomy of the condyle was often associated with signs and symptoms. We arrive at the conclusion that TMD is an umbrella, housing several overlapping problems that may manifest themselves at irregular intervals or even may disappear completely.