Multifactorial Risk Indicators for the Development of Temporomandibular Disorders: A Case-Control Study 

Caio M. P. Selaimen¹, DDS, MS; José C. M. Jeronymo¹, DDS, MS;

Diego P. Brilhante¹, DDS, MS; Eduardo M. Lima, DDS, MS, PhD¹; Márcio L. Grossi^1,2, DDS, MS, PhD 

¹Faculty of Dentistry, Pontifical Catholic University of Rio Grande do Sul - PUCRS, Brazil

²Temporomandibular Disorders Research Unit - CNPq, Brazilian Government

Address Correspondence to:

Dr. Márcio L. Grossi

PUCRS Dental School

Avenida Ipiranga 6681

Porto Alegre, RS, Brazil

CEP  90619-900

Tel: (51) 3320-3546

Fax: (51) 3320-3626

Email: mlgrossi@pucrs.br 

Abstract

Aims: This case-control study was undertaken mainly to determine the multifactorial etiology of temporomandibular disorders by analyzing occlusal, sociodemographic and neuropsychological variables as risk indicators for the development of temporomandibular disorders (TMD). Methods: Seventy-two TMD patients with myofascial pain, with or without limited opening, and arthralgia (Group I.a., I.b., and III.a. of the RDC/TMD, respectively) as well as thirty age- and sex-matched pain-free concurrent controls were included. The association between selected occlusal variables (i.e., overbite, overjet, number of anterior and posterior teeth, bilateral canine guidance on lateral and protrusive movements, anterior centric slide, Angle Classes I,II and III malocclusion), as well as sociodemographic (education, employment, income, age, race, marital status, number of children, social and physical activity, coffee and alcohol consumption) and neuropsychological (sleep and depression) variables and TMD were analyzed. Confounders were controlled in the design stage by the inclusion/exclusion criteria. Results: Spontaneous pain as well as pain on palpation (Grade II or higher) were also statistically worse in TMD patients, confirming the separation between patients and controls. Among the occlusal variables, only the presence of Class II Angle malocclusion (crude OR = 8.0, CI = 2.2-29.3) and the absence of bilateral canine guidance on lateral excursion (crude OR = 3.9, 1.6-9.7) were statistically more common in patients than controls. Significant sociodemographic as well as neuropsychological variables were sleep (the Sleep Assessment Questionnaire^© - SAQ^©), employment, age, depression (the Brazilian-Portuguese Version of the Beck Depression Inventory - BDI), cigarette and alcohol consumption. In the forward stepwise logistic regression analysis, Angle Class II malocclusion, sleep, alcohol and cigarette consumption were the best risk indicators (sensitivity = .68, specificity = .96, 82.4% percent correct). Conclusions: Some selected occlusal, sociodemographic and neuropsychological factors were shown to be important risk indicators, confirming the multifactorial nature of TMD. .

Key words: temporomandibular disorders, etiology, psychology, occlusion, orofacial pain, sleep, depression, case-control study 

Introduction

Some analytical studies have shown the role of occlusal factors as risk indicators for the development of TMD.^1,2 In a double-blind intervention study, three cohorts of healthy children and adolescents were studied over time concluding that the elimination of occlusal interferences reduced the risk of developing signs and symptoms, particularly muscle tenderness, from the third year to the end of the study.^3,4 These studies have preserved the traditional view of the biomedical model for treatment of TMD, which tries to identify only local (occlusal) causes rather than centrally-mediated pain mechanisms and reactions.^5,6 Notwithstanding, the assessment of neuropsychological functioning in TMD studies has received increased attention. The multifactorial etiology of TMD and the complete examination of all physical, emotional, and behavioral factors involved in the disease were also emphasized supporting the biopsychosocial model for chronic temporomandibular disorders (TMD).⁷ Some studies have shown that depression and somatization have been heavily implicated in chronic pain, including TMD.^8,9,10 Indeed, in one longitudinal treatment outcome study, sleep disorders and depression have been implicated as perpetuating factors in non-responding TMD patients.¹¹ In another study, the etiologic role of occlusion has been challenged in studies applying multiple logistic regression to calculate the odds ratios (relative risk) for several occlusal features.^12,13 These disagreements might be due to the fact that many studies have only assessed associations between TMD and a small number of risk factors, and there is the need of studies evaluating occlusal as well as sociodemographic and neuropsychological factors at the same time.^11,14

Therefore, the objective of this study was to analyze ten reported significant occlusal factors as well as concurrent sociodemographic and neuropsychological factors in the etiology of temporomandibular disorders in an analytical (case-control) study. ^1-4

Materials and Methods

Population, Inclusion and Exclusion Criteria

Newly diagnosed patients were selected for the study from the Pain Clinic at the Pontifical Catholic University of Rio Grande do Sul (PUCRS) Dental School, Brazil. Patients diagnosed with myofascial pain, with or without limited opening, and arthralgia (Group I.a., I.b., and III.a., Research Diagnostic Criteria for Temporomandibular Disorders) were included. In addition, only women between the ages of 15 and 60 years old were selected in order to control age and eliminate gender as confounders in the design stage.^14,15,16

            Based on the medical history, patients were excluded if they had history of muscle spasm, myositis, contracture, polyarthritis (osteoarthritis and osteoarthrosis), and acute traumatic injury, metabolic diseases (e.g., diabetes, hyperthyroidism), neurologic disorders (e.g., dyskinesia, trigeminal neuralgia), vascular disease (e.g., repetitive migraine, hypertension), neoplasia, psychiatric disorders, drug abuse, motor vehicle accidents (M.V.A.), and/or presenting with medical-dental emergencies as well as with visual, auditory, and motor impairments. In addition, patients currently receiving medication, particularly those affecting the CNS (muscle relaxant, anti-convulsive, and anti-depressant medications) or other treatments (e.g. acupuncture, physical therapy), were also excluded. Patients taking pain peripherally acting medication (analgesic and anti-inflammatory medications) were included, but a wash-out period of 3 days was required.^11,17-19

            A concurrent control group was recruited for comparison. The volunteers should have no pain (acute or chronic) complaints seeking the Faculty of Dentistry for restorative procedures.²⁰ The exclusion criteria was the same described for the TMD group, adding also no previous treatment for chronic pain conditions. This study was approved by the Institutional Review Board, and all volunteers signed a written consent form. 

History, Pain Assessment and Clinical Examination

Sociodemographic information and the pain symptom questionnaire were self-recorded by the patient under supervision of two clinical researchers. Pain intensity at rest and pain on chewing were assessed with 100mm visual analogue scales (VAS). Anchors to both scales were labeled as “no pain” and “extremely severe pain”.²¹ These measurement validity and reliability have been described elsewhere.²² In addition, other pain dimensions, such as pain duration (in months) and pain in other parts of the body were also included to compensate for the limitations of unidimensional evaluation of pain yielded by VAS.²³

To assess the occlusion and the clinical signs of TMD (clinical examination form), a third orofacial pain specialist, who did not participate in the neuropsychological testing and pain symptom questionnaire, blinded to both TMD and control groups, was chosen.²¹ The occlusal assessment was made for some occlusal variables, which have been shown to be associated with TMD^1-4: i) overbite, ii) overjet, iii) number of anterior teeth, iv) number of posterior teeth, v) Angle malocclusion classification, vi) bilateral canine guidance on lateral excursion, vii) bilateral canine guidance of protrusion, and viii) anterior centric (CR-CO) slide. Overbite, overjet, and anterior centric (CR-CO) slide were assessed using a digital caliper (Mitutoyo Digimatic Caliper, Japan), while the presence of working and balancing side interferences when the patient was in lateral excursion was assessed using a thin double-sided articulating paper (Accu Film II, Parker, Farmingdale, NY, USA). The classification of Angle malocclusion (Classes I and II) was described elsewhere;¹ Class III included cases with bilateral or unilateral mesial displacement of the lower first molar and canine of at least half a cusp.

            The examination for signs and symptoms of TMD was based on the standardized RDC/TMD (Axis I). Subjects underwent extraoral examination comprising palpation of the TMJ (lateral pole), masseter, temporalis and sternocleidomastoid muscles. Assigned scores were based on the patient’s evoked pain reaction when specific muscle and joint sites were palpated. Intraoral examination, besides the occlusal variables previously described, included evaluation of: a) maximum unassisted mandibular opening, b) percussion sensitivity, and c) presence of dental caries to exclude pain from dentoalveolar etiology.¹⁵

Clinical examination, particularly muscle/joint palpation and maximum unassisted mandibular opening measurement, was performed last; because, it has been shown to exacerbate pain and change our neuropsychological assessment.¹¹ In order to further confirm that,  muscle and/or joint pain exacerbation after examination were also checked. 

Sociodemographic and Neuropsychological Assessment   

            Some sociodemographic and clinical variables that have been shown to influence the neuropsychological tests described below, such as age, gender, language, chemical dependency, history of trauma, neurological disorders, and psychological status, were controlled for in the design stage, by restriction in the inclusion/exclusion criteria.²⁴ In an analogous way, circadian pain cycle was controlled by standardization of the day time (10 a.m. and 4 p.m.) and site (single isolated room) where the tests were performed.²⁵ Other sociodemographic variables, such as educational level, employment, income, age, race, marital status, number of children, physical activity, coffee, cigarrete and alcohol consumption were evaluated by self-completed questionnaires and assessed in the analysis stage.¹¹

            The neuropsychological assessment was only made for depression and sleep as possible risk indicators for the development of TMD.^11,26-28 In order to assess depression, the long version of the Beck Depression Inventory (scores 0-63), translated and validated into Brazilian-Portuguese language with permission (Casa do Psicológo and The Psychological Corporation, USA) was used.^29,30 The test norms, reliability and repeatability has been extensively reviewed.^29,31,32 To assess sleep, the validated (polysomnography) 19-item self-administered Sleep Assessment Questionnaire^© (SAQ)^© was used, which was translated into the Brazilian-Portuguese language with permission of the authors. Normative data, test description, reproducibility and validity have also been published in the literature.³³ The neuropsychological tests used in this study were simple and easily completed by the patient with little or no help from the single clinical psychologist supervising the neuropsychological testing, blinded to both TMD and non-pain groups, which prevented interviewer biases.³¹

 Sample Size and Data Analysis

            The formula for calculating sample size for two independent means has been described in a previous publication.^11,34 To disclose a percent difference of 40%, which we considered clinically relevant and statistically significant between the two groups (two-sided test at the 0.05 level and for a power of 80%) for all occlusal variables, it was estimated that 27 patients must be screened in each group, which was increased to 30 to compensate for drop-outs.¹⁷ Considering that a patient/control ratio of 2:1 is considered satisfactory, 60 patients and 30 controls were selected.²⁰

The Mann-Whitney U test was used for continuous variables due to data distribution; categorical variables were analyzed by two-by-two tables using Chi Square and Fisher’s Exact tests. A 0.05 significance level was adopted for all tests. Odds ratio (critical OR = 2.0) and 95% confidence interval were used to determine the strength of the association. Logistic regression was used to select the best risk indicators of TMD. ^35,36 In the logistic regression, the variables (dependent and independent) in the models had to be dichotomized to allow direct comparison. Recoding procedures (cutoff points) were based either on the normative data of the tests (i.e., SAQ^© and BDI) or, when norms were not available, we used similar dychotomization published in the literature.^{1,11,12,30,33,} Spearman’s rank correlation was used to evaluate the association among all significant variables that might be found. All analyses were performed using SPSS version 11.5 for Windows (SPSS Inc., Illinois). The TMD patients comprised Group I; and controls, Group II.

 Results

            The recruitment rate in our study was 80% among TMD patients, in one year; among controls, the recruitment rate was 100% in a period of nine months. Our combined sample (TMD and non-pain subjects) was predominantly Caucasian (72.5%), from low family income (65.4%), with some education after high school or less (18.1%), and had an average age of 34.3 years-old (± 12.8). Among occlusal variables, 58.8% were Angle Class I, 36.3% were Class II, and 4.9% were Class III. Only 7.9 % had an anterior centric slide greater than 2 mm, and 31.7% had bilateral canine guidance on lateral excursion, which was suggestive of grinding of the canine due to bruxism. 

History, Pain Assessment and Occlusal Variables

            Analysis of the clinical examination was carried out for some occlusal variables and traditional signs and symptoms of TMD.^1,2,15 Some clinical variables did not show statistically significant differences between experimental and control groups. Maximum unassisted mandibular opening in mm (Mean ± S.D.) was similar between TMD patients (50.2 ± 9.5) and non-pain subjects (47.8 ± 6.3). Likewise, overbite (2.9 ± 1.5 versus 2.8 ± 1.9) and overjet (2.9 ± 1.6 versus 2.4 ± 1.6) were similar between experimental and control groups, respectively. Mean number (± S.D.) of anterior teeth (11.6 ± 1.1 versus 11.3 ± 2.2) and posterior teeth (15.2 ± 3.1 versus 13.5 ± 5.2) were also very close between Groups I and II, respectively. Angle Class III malocclusion was present in only 7.9% of TMD patients versus 7.4% of controls, and it was not significant. Anterior centric slide (CR-CO) was also not significant; however, only 11.3% of TMD patients had anterior (CR-CO) slide greater than 2 mm versus none in the control group. Conversely, the presence of bilateral canine guidance on protrusion was unusual and less frequent in TMD than in non-pain subjects (1.4% versus 6.7%, respectively) and was also not significant. Additionally, neither group had evidence of carious lesions detectable with a probe.

On the other hand, other clinical variables did show highly statistically significant differences between TMD and non-pain subjects. The masseter, temporalis, sternocleidomastoid muscles and the temporomandibular joints – lateral pole (TMJ) palpation, following the RDC/TMD - scores II to III, was proportionately more sensitive in TMD patients than controls. Also, there was also a highly statistically significant difference between the experimental and control groups with respect to pain intensity at rest and on chewing. The TMD patients had a mild intensity pain level against no pain in the control group. In addition, 38.9% of TMD patients had exacerbation of joint and muscle pain after examination, contrasting with none of the non-pain group (Table 1). Other multidimensional measurements of pain were also relevant. The average pain duration was long in TMD patients (32.7 ± 26.6 months), and almost 70% of TMD patients reported pain in other parts of their body (e.g., headaches, back pain, abdominal pain, pain in the eye globe, general body pain, etc.).

 Regarding the significant occlusal variables, TMD patients were more likely to have absence of bilateral canine guidance on lateral movements when compared to non-pain subjects with an increase of 3.9 times the risk (OR) of being a TMD patient. Class II Angle malocclusion was also more common in the TMD group when compared to controls, with an increased risk (OR = 8.0) of developing TMD. In addition, percussion sensitivity was also more commonly found in TMD patients than in non-pain individuals (Table 1).           

Sociodemographic and Neuropsychological  Assessment

            As depicited in Table 2, sociodemographic and neuropsychological variables not controlled for in the design stage, such as educational level, employment, income, marital status, number of children, physical activity, social activity, coffee, cigarette and alcohol consumption, as well as sleep and depression, were assessed in the analysis stage (univariate and multivariate analyses).^11,24 Educational level, income, marital status, number of children, physical activity, social activity and coffee consumption did not show any statistically significant differences between TMD patients and non-pain subjects. Additionally, none of these confounders also reached our critical OR (2.0), which ranged from 0.5 to 1.7.

On the other hand, other confounders did show statistical significance (P<0.05) between patients and controls (Table 2). Age, which was controlled for in the design stage, was statistically different between experimental and control groups in the analysis stage. Unemployment increased 4.3 times the risk of being a TMD patient. Cigarette and alcohol consumption were also statistically significantly different between Groups I and II, but in both cases they acted reducing the risk of developing TMD (OR = 0.1). TMD patients showed SAQ^© sleep scores 46.6% higher than the non-pain (control) group. For the BDI, TMD patients had scores 59.2% higher than controls.    

Selection of Best Risk Indicators

In the final analysis, for the selection of our best predictors (risk indicators) of TMD, a forward stepwise analysis using the two significant occlusal as well as the sociodemographic and neuropsychological variables described above was carried out.³⁶ Angle Class II malocclusion, sleep (SAQ^©), alcohol  and cigarette consumption were considered the best predictors for the development of TMD. Canine guidance on lateral excursion, depression (BDI), employment and age were not included in the model. The overall agreement between observed and predicted cases was higher (82.46%, sensitivity = 0.68, specificity = 0.96) when all variables were included than when the variables included in the logistic regression model were entered alone, ranging from 65% to 79.3%.

To evaluate the association among all significant variables used in our model, Spearman’s rank correlation was computed (P < 0.05). Moderate to high associations between sleep and depression (r = 0.69) as well as Angle Class II malocclusion and canine guidance on lateral excursion (r = 0.35) were found. The percentage of absent bilateral canine guidance on lateral excursion was significantly more common in Class II patients than in Class I (83.3% versus 58.3%, Chi Square test, P < 0.05). Centric slide greater than 2 mm was identical between both (8.3%). The relevance of these findings will be discussed below.  

Discussion

            In this investigation, we tried to separate patients (predominantly muscle disorders) from controls in a Southern-Brazilian sample assessing selected occlusal, sociodemographic and neuropsychological variables that have been shown to be associated with TMD. A case-control design was selected due to the fact that it is the most adequate analytical study in cases where the latency period of the disease is long and/or the incidence is low.^14,20 Our recruitment rate and TMD sample size allowed the use of multivariate analysis and was similar to previous studies.^1,2,26,27,37 The primary diagnosis of myofascial pain, with or without limited opening, and arthralgia (Group I.a., I.b., and III.a. of the RDC/TMD, respectively) was chosen,¹⁵ due to the fact that it has been documented that patients diagnosed with myofascial pain and arthralgia have significantly higher levels of depression and somatization than those diagnosed with only disk displacements,¹⁰ and that TMD pain has a predominantly muscular origin (52.9%),¹⁶ which increased our external validity. This is in agreement with a current trend in the TMD literature to study well defined populations in order to increase the validity and reproducibility of the results.^{1-3,10,11,18,26,28} Indeed, studies have shown that different TMD sub-groups (e.g., post-traumatic TMD as a result of motor vehicle accident, and disk displacements and/or arthritic conditions) have different risk factors and etiology.^18,38,39  Our frequencies of Angle malocclusion and canine guidance was also similar to previous studies. ^12,13,19

            Among the occlusal variables, overbite, overjet, Angle Class III malocclusion, number of anterior and posterior teeth, and anterior centric slide did not yield significant results between TMD and non-pain groups. The only two occlusal variables found to be risk indicators for the development of TMD were bilateral canine guidance on lateral excursion and Angle Class II malocclusion (Table 1). Other studies also found that some occlusal factors, particularly Angle Class II malocclusion,¹ are important risk factors for the development of TMD.^2-4,6

            Regarding other clinical variables, palpation of TMJ (lateral pole) and masticatory (masseter and temporalis) and neck (sternocleidomastoid) muscles disclosed significant differences between TMD and controls (Table 1), confirming the separation between patients and controls for some signs and symptoms of TMD. Previous studies^18,40 also found very sensitive muscles in TMD patients, particularly non-responding and post-traumatic TMD patients. However, it must be pointed out that traditional signs and symptoms of TMD, including muscle and joint palpation, do not have a good reproducibility and were not considered good predictors of treatment outcome.^11,21,41 In addition, it has been shown that dysfunctional (non-responding) TMD patients usually present positive responses to placebo sites.⁴² Maximum unassisted mandibular opening was neither significant nor relevant between the two groups; and only 6% of our TMD and none in our controls had incisal opening less than 35 mm, which agrees with previous literature where the range has been reported from 0% to 5%.^11,18,40

            The pain levels at rest (100 mm VAS) for the TMD group in this investigation was of medium intensity and comparable to previous studies (35 to 64 mm).^11,17,43 Our pain on chewing was gently higher, and this aggravation of pain during function was also reported by Grossi et al. and Dao et al. ^11,17 The former found it a good predictor of TMD treatment outcome. Pain exacerbation was positive for the TMD group, but not for controls, which justified our choice of performing the neuropsychological tests prior to clinical examination; otherwise, this increase in pain levels would have most likely affected our test results (Table 1). The average pain duration was long (32.7 months), which was similar to the average described in the literature (41.6 to 72 months) and confirmed the chronic nature of TMD.^11,43-45 In this study’s sample, 68.1% of TMD patients reported pain in other parts of the body against none in the control group (P < 0.001), including back pain, general muscle pain, headaches and abdominal pain. Similar studies also reported high levels of comorbidity between TMD with fibromylgia and irritable bowel syndrome.^11,46

            Regarding sociodemographic variables, all TMD subjects in this study were females in order to increase the internal validity and to control an important confounder in the design stage.²⁰ Despite that, the gender distribution seems to be comparable to other studies, 65% to 100%, considering that the percentage of women has been much greater than men, also granting good external validity to the results.^{1,26,27,37,38} Differently than previous studies,^11,26,27,37 the majority of our TMD population did not have post-secondary diploma certificate or higher education as well as high levels of unemployment (Table 2). Similar to our previous study (57.9%), this study’s sample also had predominantly low income level; but it contrasted (26.4%) with another one.^11,37 Unfortunately, not all studies reported employment and income level, and it was difficult to assess their overall differences. The average age of the TMD sample in this investigation was similar to what is found in the literature (mean = 27.4 to 47, range = 5 to 61),^10,26,37,39 but it was significantly lower than our control group (Table 2). In general, the social and demographic variables in this study sample seem to be comparable with similar studies in the literature.

            Regarding neuropsychological confounders, the levels of depression, as measured by the Beck Depression Inventory (BDI),^29,30 was significantly higher in TMD patients than in controls (Table 2). These BDI scores for TMD patients and controls were comparable with previous studies.^11,29,30 They were also in agreement with others who used different methods to measure the prevalence of depression in both TMD (15% to 43%)^18,37,39,40 and non-pain subjects (6%).⁴⁷ Sleep, as measured by the Sleep Assessment Questionnaire^© (SAQ^©) was an important risk indicator in the development of TMD. The SAQ^© scores for TMD patients were significantly higher than for controls, confirming our previous findings¹¹ and others who did not use the SAQ^© but found close association between sleep and TMD or other chronic pains.^26,48-50

In our final analysis, for the selection of our best predictors (risk indicators) of TMD, a forward stepwise analysis using the significant occlusal, sociodemographic and neuropsychological variables was carried out.³⁶ Angle Class II malocclusion, sleep (SAQ^©), alcohol  and cigarette consumption were considered the best predictors for the development of TMD. Canine guidance on lateral excursion, depression (BDI), employment and age were not included in the model. The overall agreement was higher when all variables were included in the model than when the variables were placed alone; demonstrating that the combination of selected occlusal and neuropsychological variables is a better predictor than when the variables are placed alone.

Assessing the correlation among the eight variables included in our model, we found moderate to high associations between sleep and depression as well as Angle Class II malocclusion and canine guidance on lateral excursion. This helps to explain why sleep was included as one of the best TMD predictors, while depression was not. In order to clarify the association between Angle Class II and bilateral canine guidance on lateral excursion, the percentage of canine guidance among Angle Class I, II and III was calculated. The percentage of absence of bilateral canine guidance on lateral excursion was higher in Class II patients than in Class I. These data suggested that the association of Class II with TMD was partially due to the high proportion of absence bilateral canine guidance on lateral excursion. These results were very similar to previous studies where the association between Angle Class II and TMD was found.^1,51 The overall results confirmed the multifactorial etiology of TMD and the importance of their multidisciplinary management, supporting the biopsychosocial model of chronic pain.  

Bias and suggestions for future studies

            Despite our efforts to blind and standardize our clinical, sociodemographic and neuropsychological evaluation, case-control studies are always vulnerable to biases and the associations presented here cannot imply cause-and-effect relationships between TMD and any variable studied here.²⁰ Therefore, it is necessary to confirm our findings with analytical longitudinal designs. Also, considering that we found clinical indications of presence of bruxism, such as percussion sensitivity, high prevalence of sleep disorders, and high prevalence of absence of bilateral canine guidance, both in lateral and protrusive movements, it is important to further assess the role of bruxism, using polysomnographic evaluation, in the initiation and perpetuation of TMD. Indeed, two studies found that related symptoms, such as tooth clenching, are important risk factors in TMD.^1,14,48 The rationale for the association between Angle Class II and TMD is still debatable and needs further verification.⁵² 

Acknowledgements 

            The authors wish to thank Dr. Harvey Moldofsky, who gave us the permission of translating and testing the English-version of the Sleep Assessment Questionnaire^© (SAQ^©) into Brazilian-Portuguese, which was essential in providing the adequate methodology for this manuscript. We also whish to extend our gratitute to Dr. Nilton Sodi Saueressig and the clinical psychologist Maria Rosana de Oliveira, for their assistance in patient recruitment and supervision in the neuropsychological testing. Finally, we would also like to thank Dr. John Rugh and Dr. Herenia Lawrence for reviewing this manuscript. 

References 

1.    Henrikson T, Ekberg EC, Nilner M. Symptoms and signs of temporomandibular disorders in girls with normal occlusion and class II malocclusion. Acta Odontol Scand 1997;55:229-235.

2.    Karjalainen M, Bell YL, Jämsä T, Karjalainen S. Prevention of temporomandibular disorder-related signs and symptoms in orthodontically treated adolescents. Acta Odontol Scand 1997;55:319-324.

3.    Kirveskari P, Alanen P, Jämsä T. Association between craniomandibular disorder and occlusal interferences in children. J Prosthet Dent 1992;67:692-696.

4.    Kirveskari P. The role of occlusal adjustment in the management of temporomandibular disorders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997;83:87-90.

5.    Dawson, P. Position paper regarding diagnosis, management and treatment of temporomandibular disorders. The Equilibration Society. J Prosthet Dent 1999;81:174-178.

6.    Kirveskari P, Alanen P, Jämsä T. Association between craniomandibular disorder and occlusal interferences. J Prosthet Dent 1989;62:66-69.

7.    Rugh JD: Psychologic components of pain. Dent Clin North Am 1987;31:579-594.Beck AT: Depression: Causes and Treatment. 1st. ed. Philadelphia, University of Pennsylvania Press, 1970.

8.    Dworkin SF, Von Korff MR, LeResche L Multiple pains and psychiatric disturbance: an epidemiologic investigation. Arch Gen Psychiatry 1990:47;239-44.

9.    McCreary CP, Clark GT, Oakley ME, Flack V: Predicting response to treatment for temporomandibular disorders. J Craniomand Disord Facial Oral Pain 1992;6:161-70.

10.Yap AUJ, Tan KBC, Chua EK, Tan HH. Depression and somatization in patients with temporomandibular disorders. J Prosthet Dent 2002;88:479-84.

11.Grossi ML, Goldberg MB, Locker D, Tenenbaum HC. Reduced Neuropsychologic Measures as Predictors of Treatment Outcome in Patients with Temporomandibular Disorders. J Orofac Pain 2001;15:329-339.

12.Pullinger AG, Seligman DA, Gornbein JA. A multiple logistic regression analysis of the risk and relative odds of temporomandibular disorders as a function of common occlusal features. J Dent Res 1993;72:968-79.

13.Pullinger AG, Seligman DA. Quantification and validation of predictive values of occlusal variables in temporomandibular disorders using a multifactorial analysis. J Prosthet Dent 2000;83:66-75.

14.Velly AM, Gornitsky M, Philippe P. Contributing factors to chronic myofascial pain: a case-control study. Pain 2003;104:491-499.

15.Dworkin SF, LeResche L, eds. Research Diagnostic Criteria for Temporomandibular Disorders. J Craniomandibular Disord Facial Oral Pain 1992; 6(4):301-355.

16.Stohler CS. Clinical Perspectives on Masticatory and Related Muscle Disorders. In: Temporomandibular Disorders and Related Pain Conditions. Progress in Pain Research and Management, Vol. 4, edited by B.J. Sessle, P.S. Bryant, and R.A. Dionne, IASP Press, Seattle, 1995.

17.Dao TTT, Lavigne GJ, Charbonneau A, Feine JS, Lund, JP. The efficacy of oral splints in the treatment of myofascial pain of the jaw muscles: a controlled clinical trial. Pain 1994;56:85-94.

18.Goldberg MB, Mock, D, Ichise M, Proulx G, Gordon A, Shandling M, Tsai S, Tenenbaum HC. Neuropsychologic Deficits and Clinical Features of Posttraumatic Temporomandibular Disorders. J Orofac Pain 1996;10(2):126-140.

19. Ferraro VF, Sforza C, Miani A, Serrao G. Kinesiographic three-dimensional evaluation of mandibular border movements: a statistical study in a normal young nonpatient group. J Prosthet Dent. 1992 Oct;68(4):672-6.

20.Hennekens CH, Buring JE. Epidemiology in Medicine. Little, Brown and Company, Boston/Toronto, 1987

21.Carlsson GE, Egermark-Eriksson I, Magnusson T: Intra- and inter-observer variation in functional examination of the masticatory system. Swed Dent J 1980;4:187-194.

22.Price DD, Bush FM, Long S, Harkins SW: A comparison of pain measurement characteristics of mechanical visual analogue and simple numerical rating scales. Pain 1994;56:217-226.

23.Magnusson T, List T and Helkimo M: Self-assessment of pain and discomfort in patients with temporomandibular disorders: a comparison of five different scales with respect to their precision and sensitivity as well as their capacity to register memory of pain and discomfort. J Oral Rehab 1995;22:549-556.

24.Stuss DT, Levine B. The dementias: Nosological and Clinical Factors related to Diagnosis. Rotman Research Institute of Baycrest Centre, University of Toronto, 1995.

25.Folkard S: Diurnal variation and individual differences in the perception of intractable pain. J Psychosomatic Research 1976;20:289-301.

26.Fricton JR, Olsen T. Predictors of outcome for treatment of temporomandibular disorders. J Orofacial Pain 1996; 10:54-65.

27.Lipton JA, Marbach JJ: Predictors of treatment outcome in patients with myofascial pain-dysfunction syndrome and organic temporomandibular joint disorders. J Prosth Dent 1984;51(3):387-393.

28.Vallon D, Nilner M, Soderfeldt B. Treatment outcome in patients with craniomandibular disorders of muscular origin: a 7-year follow-up. J Orofac Pain 1998; 12:210-8.

29.Beck AT. Depression: Causes and Treatment. 1st. ed. Philadelphia, University of Pennsylvania Press, 1970.

30.Cunha JA, Prieb RGG, Goulart PM, Lemes RB. The Use of the Beck Depression Inventory to Evaluate Depression in University Students. Psico 1996;27(1):107-115. (Published in Portuguese).

31.Mapou RL, Spector J. Clinical Neuropsychological Assessment: A Cognitive Approach. New York and London, Plenum Press, 1995.

32.Schwartz JE, Jandorf L, Krupp LB. The measurement of fatigue: a new instrument. J Psychosomatic Res 1993;37:753-762.

33.Cesta A, Moldofsky H, Sammut C: The University of Toronto Sleep Assessment Questionnaire^© (SAQ^©). Sleep Research 1996;25:486.

34.Taylor W: How to use guide to sample size calculation. McMaster University, 1981.

35.Greenland S: Modeling and Variable Selection in Epidemiologic Analysis. AJPH 1989; 79(3):340-349.

36.Norman GR, Streiner DL. PDQ Statistics. 2nd Edition, Mosby, 1997.

37.Gerschman JA, Wright JL, Hall WD, Reade PC, Burrows GD, Holwill BJ. Comparisons of psychological and social factors in patients with chronic orofacial pain and dental phobic disorders. Aust Dent J 1987; 32:331-335.

38.Pullinger AG, Seligman DA, John MT, Harkins S. Multifactorial comparison of disk displacement with and without reduction to normals according to temporomandibular joint hard tissue anatomic relationships. J Prosthet Dent 2002;87:298-310.

39.Yap AUJ, Dworkin SF, Chua EK, List t, Tan KBC, Tan HH. Prevalence of Temporomandibular Disorder Subtypes, Psychologic Distress, and Psychosocial Dysfunction in Asian Patients. J Orofac Pain 2003;17:21-28.

40.Ohrbach R, Dworkin SF. Five-year outcomes in TMD: relationship of changes in pain to changes in physical and psychological variables. Pain 1998 Feb;74(2-3):315-26.

41.DeWijer A, Lobbezoo-Scholte AM, Steenks MH, Bosman F: Reliability of Clinical Findings in Temporomandibular Disorders. J Orofac Pain 1995;9:181-191.

42.Wilson L, Dworkin SF, Whitney CW, LeResche L: Somatization and pain dispersion in chronic temporomandibular disorder pain. Pain 1994;57:55-61.

43.Linton SJ, Melin L: The Accuracy of Remembering Chronic Pain. Pain 1982;13:281-285.

44.Gale EN, Funch DP: Factors associated with successful outcome from behavioral therapy for chronic temporomandibular joint pain. J Psychosom Res 1984;28(6):441-448.

45.Linton SJ, Gotestam KG: A Clinical Comparison of Two Pain Scales: Correlation, Remembering Chronic Pain, and a Measure of Compliance. Pain 1983;17:57-65.

46.Dao TTT, Reynolds WJ, Tenenbaum HC. Comorbidity between Myofascial Pain of the Masticatory Muscles and Fibromyalgia. J Orofac Pain 1997 Summer;11(3):232-41.

47.Nielsen AC, Williams TA: Depression in ambulatory medical patients: prevalence by self-report questionnaire and recognition by nonpsychiatric physicians. Arch Gen Psychiatry 1980; 37:999-1004.

48.Lavigne GJ, Kato T, Kolta A, Sessle BJ. Neurobiological Mechanisms Involved in Sleep Bruxism. Crit Rev Oral Biol Med 2003;14(1):30-46.

49.Lentz MJ, Landis CA, Rohermel J, Shaver JL. Effects of selective slow wave sleep disruption of musculoskeletal pain and fatigue in middle age women. J Rheumatol 1999;26(7):1586-1592.

50.Unger ER, Nisenbaum R, Moldofsky H, Cesta A, Sammut C, Reyes M, et al. Sleep assessment in a population-based study of chronic fatigue syndrome. BMC Neurology 2004 Apr 19; 4(1):6.

51.Henrikson T, Nilner M. Temporomandibular disorders, occlusion and orthodontic treatment. J of Orthod 2003;30:129-137.

52.Gaggl A, Schultes G, Santler G, Kärcher H, Simbrunner J. Clinical and magnetic resonance findings in the temporomandibular joints of patients before and after orthognathic surgery. Brit J Oral Maxillofac Surg 1999;37:41-45. 

Table 1

Significant Signs and Symptoms and Occlusal Variables in TMD (Group I) versus

Non-pain (Group II) Subjects

* P < 0.05, † P < 0.01, ‡ P < 0.001.

§ Critical Odds Ratio = 2.0.

NA = non-available (not risk indicators)

Table 2

Sociodemographic and Neuropsychological  Variables in TMD (Group I) versus Non-pain (Group II) Subjects