Adequate data on the incidence, prevalence, natural history, and clinical course of temperomandibular disorders (TMD) and other chronic pain conditions are largely lacking, though the need to derive such basic data is recognized by clinicians, researchers, and public health agencies. This paper discusses challenges to the epidemiologic study of TMD diagnosis. These challenges include:
• Case definition: There is currently poor agreement regarding which combinations of clinical and psychosocial findings differentially define cases of TMD
• Differentiation of normal variation v pathophysiologic signs: To what extent do commonly gathered clinical measurements constitute pathophysiologic signs of TMD v reflect normal biologic variation
• Reliability of clinical measurement: Factors influencing reliability of clinical signs and reliability of examiners have not been adequately assessed
• Progressive v self-limiting disease activity: Do TMD subtypes represent a continuum of pathologic disease activity, or nonmutually exclusive categories describing largely symptomatic pain conditions that are selflimiting or stable.
It is recommended that epidemiologic studies not be constrained by a priori definitions of TMD subtypes, but continue to gather data on clinical signs and symptoms that have theoretical and clinical relevance to mandibular dysfunction and psychosocial status. An approach is proposed for development of reliable and valid criteria of TMD subtypes suitable for epidemiologic research.