Some studies also have suggested that parafunctions are associated with TMD pain. For example, a survey of 483 adults living in northern Italy reported a relationship between parafunctional bruxism and craniofacial pain,⁷ and similar findings have been reported by other researchers.^8–10 Experimental approaches confirm that parafunctions can increase pain and lead to a short-term diagnosis of TMD in otherwise healthy people. Pain-free subjects in one study who were instructed to bite with maximum force reported increased pain after only a few minutes of such activity.¹¹ Considerably less intense clenching carried out for longer periods^12–14 also can result in pain that meets the standards for TMD, as defined by the research diagnostic criteria for TMD (RDC/TMD) set forth by Dworkin and LeResche.¹⁵

Additionally, people who suffer from chronic pain frequently report a number of emotional changes. In patients with TMD, one common correlate of chronic pain is alteration in mood, typically depression. Gatchel and colleagues¹⁶ showed that patients with chronic TMD were more likely to have significant signs of depression than were patients with more recent onset of TMD. Psychological disturbances are not uniformly distributed among patients with TMD, but they are more common in those with chronic muscular pain than in those with nonpainful disk and joint problems.^17,18

Studies that use traditional surveys to assess the role of parafunctions and stress in TMD pain may suffer from a number of methodological problems. Primary among these is the retrospective nature of data that are based on subjects’ recall. A better strategy is to use experience sampling methodology (ESM) (also known as "ecological momentary assessment"[EMA]) to examine the role of stress and parafunctions in patients with TMD. ESM is characterized by repeated measurements in a natural environment. The high level of repeated sampling in ESM gives an estimate of the true value of the phenomenon in the natural environment.¹⁹ In addition, the studies reviewed here typically examined group differences in mood, parafunctions and stress and did not attempt to predict pain levels reported by individual subjects.

We undertook a study of TMD pain, parafunctions, mood and stress, using ESM as a technique for assessing these factors. The hypothesis of this study was that parafunctions, mood and stress contributed individually to facial and jaw pain reported by people without TMD, people who have TMD but no pain, and people who have TMD accompanied by pain.

	SUBJECTS AND METHODS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Subjects. We selected subjects (initial N = 113) from among patients of the University of Missouri-Kansas City Facial Pain Center or recruited them from the general population via newspaper and radio advertisements. Exclusion criteria for the study included any history of major trauma to the head or neck; current use of an intraoral appliance; active orthodontic treatment; history of any chronic pain condition besides TMD; or current daily use of any analgesic, antidepressant or muscle relaxant medication. Ten people dropped out of the study before completing it.

Screening examination. Researchers assessed all subjects using the RDC/TMD.¹⁵ Each person was assessed by two independent examiners (usually, but not exclusively, A.G.G. and L.L.), and each examiner was blind to the diagnosis provided by the other. The time between examinations was approximately 30 minutes. Agreement between raters regarding whether a specific condition was present ranged between 64.3 percent for arthralgia and 88.9 percent for myofascial pain.²⁰ The presence of disk displacement was inconsistent in three subjects, and raters therefore failed to agree on a diagnosis for these three people.

The examining clinicians palpated 16 muscle sites accessible extraorally (left and right anterior temporalis, middle temporalis, posterior temporalis, origin of masseter, body of masseter, insertion of masseter, posterior mandibular region and submandibular region) and four muscle sites accessible intraorally (left and right temporalis tendon and the lateral pterygoid area), according to the techniques described in the RDC. A subject scored his or her pain during muscle palpation on a 0-to-3 scale, with 0 signifying "no pain." The examining clinicians determined the presence of reproducible clicking on vertical opening, closing, lateral excursion and protrusion by means of palpation, supplemented by auscultation, as they did with the presence of coarse crepitus. Subjects rated pain in the temporomandibular joint (TMJ) produced by palpation and rated it on a 0-to-3 scale. Examiners measured pain-free unassisted mandibular opening and maximum unassisted opening in millimeters. Only the potential subjects who received the same diagnosis from the raters were eligible to participate in the study. We excluded from participation any prospective subject who received a diagnosis of osteoarthritis or osteoarthrosis of the TMJ according to RDC criteria.

On the basis of the results of the screening examination and subjects’ self-reported data, we assigned subjects who qualified for participation to one of four groups based on RDC criteria:

– myofascial pain (M);

– myofascial pain and arthralgia (MA);

– disk displacement (DD);

– control subjects without TMD.

Within each of the three TMD groups, subjects received only the diagnosis or diagnoses contained within the group’s title and no other RDC diagnosis. By definition, two of the groups had significant pain in the face and jaw area and two did not.

Sampling method. To obtain more accurate measures of pain and other states, we used ESM. Subjects in this study carried pagers, one-way devices that beeped or vibrated when contacted. A custom-programmed executable binary file derived from the Paradox 7 database (Corel, Ottawa) placed calls to pagers. The mean time between calls was 120 minutes, with a 40-minute window of variability within which a specific call could be placed; a specific call to a subject could occur up to 20 minutes earlier or up to 20 minutes later than would be expected on a fixed, invariant schedule. The variability of calls was based on a random-number generator that produced an equal distribution of values on either side of the expected call time. Variability in calling schedules reduced the possibility that subjects’ behavior would be affected by the anticipation of a call at a fixed point in time.

Subjects were asked to report on pain in the jaw, face or head; the presence and intensity of tooth contact; tension in the jaw, face or head; mood; and stress level.

A research assistant instructed subjects to fill out a 3 x 5-inch card on which was printed a brief questionnaire each time they were paged, unless doing so would jeopardize their safety. Subjects were asked to report on pain in the jaw, face or head; the presence and intensity of tooth contact; tension in the jaw, face or head; mood; and stress level. Except for tooth contact, all measures used an 11-point (0–10) numerical rating scale. The anchors for the pain measures were "no pain" and "severe pain," the anchors for the tension measure were "completely relaxed" and "extremely tense," and the anchors for the two mood measures were "irritable" and "cheerful" for one scale (Mood 1) and "happy" and "sad" for the other (Mood 2). We reverse-scored the two mood scales as a measure of response validity. The anchors for the stress scale were "no stress" and "extremely high stress." Tooth contact was scored on a 4-point scale, ranging from "no contact" to "strong clenching." We instructed subjects that they could turn off the pager if they went to bed early or stayed in bed late or were in an environment in which paging would be dangerous or disruptive.

Procedure. After obtaining informed consent, a research assistant blinded to the subject’s diagnosis was introduced to the subject. Following a script, the assistant instructed the subject on the use of the pager. Each subject had multiple opportunities to interact with the device and to demonstrate competence in turning on the device, responding to a page, changing the battery and filling out the questionnaire. The research assistant asked subjects at what time each day they would become fully alert and capable of responding to a page after awakening, and also at what time they typically retired for the evening. This allowed for individual flexibility in generating call schedules appropriate for each subject.

The first day of paging for each subject varied randomly from Monday through Sunday and continued for one week. The times at which subjects were willing to receive a page in the morning and were no longer willing to receive a page in the evening were biased toward times on the hour (for example, 7 a.m.) or on the half-hour (for example, 10:30 p.m.). We randomized the subjects’ requested times by up to 15 minutes on either side of the requested time before entering them into the dialer program. We did not contact subjects during their normal sleep hours.

Statistical analysis. We calculated descriptive statistics by subject for each dependent variable. Because the two mood items were reverse-scored, we expected that subjects would show a negative correlation between the two items. We took the presence of a positive correlation between the two mood items as a measure of a subject’s invalid response set, and we removed these people from the subject pool (n = 4) before conducting further data analysis. We also created a weighted composite "effort" variable to estimate the combined impact of clenching intensity and time. We carried out analyses on the data provided by the remaining 96 subjects.

To assess differences between groups regarding the variables used in the study, we analyzed data by means of a one-way analysis of variance. We performed post-hoc analysis using the Student-Newman-Keuls test.

Linear regression analyses were used to predict jaw pain in this sample. Predictor variables for the regression model included tension, mood, stress and the composite "effort" variable. All predictor variables were entered into the model simultaneously. The two mood variables were highly correlated (–0.865), and we selected for the regression analysis the one measure that accounted for the greater explained variance in the groups (Mood 2). For all analyses, = .05.

	RESULTS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Table 1 presents the demographic characteristics of the four groups. There were no significant differences among groups with respect to age, education or sex. The mean number of response cards completed by each subject was 47.7 (standard deviation [SD] = 9.7).

We also carried out a third set of analyses, using only patients who had a TMD diagnosis. The model including all four predictor variables was significant, F(4,64) = 33.07, P < .001, adjusted R² = .65, as was the model without muscle tension as a predictor, F(3,65) = 17.85, P < .001, adjusted R² = .43. As indicated by the data in Table 4, these results were similar to those of the analyses that included control subjects.

	DISCUSSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

The predictors entered into the initial model accounted for 69 percent of the overall variability in jaw pain ratings. In this model, muscle tension in the jaw, face, head or a combination of these was a significant predictor of facial and jaw pain. When entered into the regression model, it accounted for a significant proportion of the variance in jaw pain, and the remaining predictors accounted for very little of the unique variance in jaw, facial and head pain. These data suggest that muscle tension, whether in the form of parafunctional tooth contact, "setting" of the jaw or other daytime parafunctions is strongly and meaningfully related to jaw pain.

It is possible that nocturnal parafunctional activity (such as tooth grinding or clenching) also contributed to reports of jaw pain, but available evidence on this point is mixed. Some investigators have concluded that the link is weak,^23–25 while others have reported a dose-response relationship between severity of parafunctions and pain in the masticatory muscles and TMJ.²⁶ The methodology used in this study investigated the impact of daytime states and behaviors on jaw pain and precludes us from making any statements about the relationship of pain to nocturnal events.

The methodology used in this study precludes any statements about the relationship of pain to nocturnal events.

In the initial model, the potential relationship between effort, mood and stress and the criterion of jaw pain may have been obscured by the inter-correlation among these predictors, as well as between the predictors and tension. Therefore, we tested a second regression model by removing muscle tension as a predictor of jaw pain and retaining the other three predictors. Overall, this model accounted for 46.2 percent of the overall variability in jaw pain scores. This model explains a smaller proportion of the variance than does the initial model, but the proportion of variance accounted for (46.2 percent) still is considerable. In this model, the coefficient for effort—a composite variable formed from the amount of time that patients reported parafunctional contact with the reported intensity of contact—was significant, as were the coefficients for mood and stress. Greater effort, more sadness and increased stress²⁷ were associated with more pain.

The results from the regression analyses were remarkably similar, regardless of whether control subjects were included. This suggests that the findings were not mathematically dependent on the presence of the control group. It also suggests that a common set of variables is associated with pain in all of the subjects with TMD who participated in the study.

The exclusion criteria used in this study ensured that the results would not be contaminated by extraneous factors such as use of intraoral appliances or analgesics. At the same time, the exclusion criteria also probably eliminated subjects who had TMD along with significant psychological disturbances and disabilities. It is likely that a patient initially seeking help for TMD-related pain will not yet have developed severe psychological complications as a result of this chronic pain condition. Therefore, the findings from this study may be especially applicable to clinicians in general practice.

The control group did not meet the criteria for TMD and did not report any chronic pain problems. This does not mean that control subjects were completely free of pain. Otherwise healthy people can have an occasional headache and other pains, and the methodology used here allowed us to capture this possibility. On an 11-point scale used to assess pain, the 0.39 value reported by control subjects indicates a very small degree of pain.

The classic self-report measures of emotional disturbance (such as the Beck Depression Inventory [BDI] and the Minnesota Multiphasic Personality Inventory [MMPI]) take a considerable amount of time to administer and score. The BDI, for example, contains 21 items, and the MMPI contains 567 items. Obviously, measures such as these cannot be used to assess emotional states quickly and repeatedly. In a pilot study, we assessed the validity of the measures we used in this investigation (A.G.G., unpublished data, October 1999). We conducted a principal-components analysis of the items and found that items had clustered together in the predicted ways. We are reasonably confident that the measures we used adequately captured the information we hoped to obtain.

The essentially correlational nature of regression analyses makes it impossible to draw cause-effect conclusions about the impact of parafunctions and psychological distress on pain. However, the findings of this study suggest that behavioral and psychological factors may contribute independently to reports of TMD-related pain.²⁸ The results of this study are consistent with the stress-muscular hyperactivity model of TMD, and they also may be consistent with other models of TMD. High levels of muscle tension and parafunctional activity may have multiple biological consequences, including microtrauma to the joint and muscles, increased levels of proinflammatory cytokines in the joint, and sensitization of pain pathways.^29–34 An alternative hypothesis is that parafunctions are an adaptive response to stress.³⁵

	CONCLUSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Our findings suggest that clinicians should carefully assess and provide guidance for reducing muscle tension, parafunctional activity, emotional distress and high levels of overall stress in their patients experiencing TMD-related myofascial pain. Studies that have used behavioral and psychological techniques specifically aimed at reducing parafunctions have reported success in reducing pain.^36–38 In contrast, there is little evidence that tranquilizers reduce pain.³⁹ Botulinum toxin injections may have short-term efficacy,⁴⁰ but the long-term effects as assessed in well-controlled double-blind trials are unknown. Clinicians who typically use intraoral appliances to treat pain in patients with TMD may suggest to their patients that appliances can act as behavior-change devices, reminding patients to relax their masticatory muscles and to reduce parafunctional activity.⁴¹ In all cases, patients will benefit from the care of clinicians who assess both biological and behavioral/psychological factors and use both sets of data in generating a treatment plan for patients with TMD who report having chronic pain.⁴²