As the title indicates, "Clinical Implications of Sex in Acute Temporomandibular Disorders" by Drs. Phillips, Gatchel, Wesley and Ellis (January JADA) addresses several important topics. I wish to comment on three: 1) how subjects were designated as either acute or chronic; 2) whether grouping by sex was a proxy for grouping by temporomandibular disorder, or TMD, subtype; and 3) how subjects were diagnosed as to disorder.
The authors state, "We classified patients as having acute TMD if they had never been diagnosed as having TMD or had been diagnosed less than six months before this evaluation. This was determined to be the most objective and accurate method of classification, since memory of past symptoms is often inaccurate." According to the authors, one can, for example, experience symptoms for anywhere from one month to 30-plus years and still meet criteria for the acute category, as long as one remains undiagnosed with TMD.
This definition of "acute" differs from that used by the community of pain clinicians and researchers.1 The customary cut point between acute and chronic pain is a report of symptom onset within the past six months.
Although the authors state that recall of pain onset is inaccurate, research on recall of facial pain onset suggests otherwise: "... high intraclass [ICC] correlations indicate that patients’ dating of pain onset may have satisfactory reliability for research purposes, when comparing a group of patients to one another to assess relative chronicity of pain."2 This latter study showed that 125 myofascial face pain patients could, in aggregate, recall pain onset ranging from one to 22 years (average seven years) reliably (ICC = .80 for the full sample).
Lack of comparability between the authors and the pain community in defining "acute" pain seriously hinders interpretation of their conclusion regarding acute vs. chronic groups.
The authors then make another critical departure regarding diagnostic issues. As is often the case in TMD literature, they begin by acknowledging that the term TMD refers to a heterogeneous group of disorders and end by concluding that "significant differences exist between men and women in regard to acute TMD symptoms."
They apply the Research Diagnostic Criteria, or RDC, for temporomandibular disorders,3 which recognizes three groups: Axis I Group 1 disorders (myofascial type TMD); Axis I Group 2 disorders (disc dysfunction); and Axis I Group 3 disorders (arthritis, and so forth).
Try as I might, I was unable to determine the number or percentage of men and women in each of the three TMD groups. Furthermore, we could find no explanation for why TMD Group 2 was omitted in the results. The latter may have contributed to the fact that proportions of each TMD group, broken down by sex and chronicity, in no category totaled 100 percent (see Table 2).
Herein lies the second major problem. As all acknowledge, TMD is a heterogeneous group of disorders. Thus, for example, it would be uninformative to compute the average age of TMD patients because the subtypes have such different age distributions. Group 1 typically has a mean of about 38 years, Group 2 occurs at almost all ages, and Group 3 usually affects people older than 50 years.
For the same reasons, one cannot speak meaningfully of the incidence, prevalence, sex ratio or most any characteristic of TMD. Moreover, specifically regarding classifying patients as acute or chronic, the same problem of heterogeneity persists. The symptoms of Groups 1, 2 and 3 are all characterized by different temporal patterns of onsets and offsets.4–6 In the absence of the ability to calculate onset of TMD, determining chronicity is impossible.
TMD is not analogous to breast cancer, in which the prevalence among women and men is markedly different while the histology appears to be the same. Reporting TMD male/female differences in psychological factors would be comparable to reporting sex differences in the rates of "pancreatic disease" when the category includes both pancreatic cancer and diabetes. Any difference noted by the authors could be the result of differences in the prevalence of the different disorders by sex and not sex itself.
It confuses the construct of cause to state that "some treatments may be more beneficial for women than for men," when it is likely more accurate to state that some treatments may be more beneficial for some TMD subtypes than for others, and that these subtypes are distributed differently by sex. As is increasingly acknowledged by researchers when examining racial factors in distribution of health outcomes, it is important to clarify whether a demographic classification variable is a proxy for a social factor, an indicator of culture or a biological variate.
These two issues comprise the major conceptual problems in the article. Additional methodological and statistical problems further compromise any conclusions that can be made, including presenting no evidence for the reliability of either TMD or psychiatric examination procedures. This is especially important given that a clinical psychology graduate student conducted both the TMD diagnostic examinations and, apparently, the psychiatric examination. Whether the examination results would have been comparable to results of calibrated clinicians is unknown. Readers are urged to view the conclusions presented in this article with great caution.
The reader should not view me as an unrepentant curmudgeon. I do, in fact, strongly endorse one of the conclusions found in the article. The authors state that "this study highlights the fact that men and women are not homogeneous, which has important implications for future research, as well as clinical applications." Vive la différence.
TOP
REFERENCES
