| Sign In to gain access to subscriptions and/or personal tools. |
Influence of Psychological Factors on Risk of Temporomandibular Disorders
1 Australian Research Centre for Population Oral Health, Dental School, University of Adelaide, SA 5005, Australia; Correspondence: * corresponding author, gary.slade{at}adelaide.edu.au
Psychological characteristics potentially may be a cause or consequence of temporomandibular disorder (TMD). We hypothesized that psychological characteristics associated with pain sensitivity would influence risk of first-onset TMD, but the effect could be attributed to variation in the gene encoding catechol-O-methyltransferase (COMT). We undertook a prospective cohort study of healthy female volunteers aged 18–34 yrs. At baseline, participants were genotyped, they completed psychological questionnaires, and underwent quantitative sensory testing to determine pain sensitivity. We followed 171 participants for up to three years, and 8.8% of them were diagnosed with first-onset TMD. Depression, perceived stress, and mood were associated with pain sensitivity and were predictive of 2- to 3-fold increases in risk of TMD (P < 0.05). However, the magnitude of increased TMD risk due to psychological factors remained unchanged after adjustment for the COMT haplotype. Psychological factors linked to pain sensitivity influenced TMD risk independently of the effects of the COMT haplotype on TMD risk.
Key Words: temporomandibular joint disorders • cohort studies • psychology • catechol O-methyltransferase • genetics
It is an enigma that individuals can vary considerably in their sensory and emotional responses to clinical conditions that have apparently similar potential to cause pain. This phenomenon is observed frequently in temporomandibular disorders (TMD), a group of chronic orofacial pain conditions that affect approximately 10% of adults in developed nations (Dworkin and LeResche, 1992; Okeson, 1996). Human experimental studies confirm extensive inter-individual variability in reported levels of pain elicited by standardized noxious stimuli (Gracely, 1999), suggesting that genetic and environmental factors influence pain perception. TMD, in turn, appears to be related to a general state of pain amplification (Maixner et al., 1998). A systematic review of epidemiological studies found that pre-existing pain conditions, depression, and female sex were consistent risk factors for TMD (Drangsholt and LeResche, 1999). Associations between psychological characteristics and pain potentially could be explained by variations in people’s genetic characteristics. Variation in a single nucleotide polymorphism (SNP) of the gene encoding catechol-O-methyltransferase (COMT) is both predictive of anxiety (Enoch et al., 2003) and associated with pain regulation in the central nervous system (Zubieta et al., 2003). The same SNP has been implicated as a risk factor for schizophrenia (Egan et al., 2001), although a systematic review found the evidence to be inconsistent (Glatt et al., 2003). However, when Shifman et al.(2002) analyzed a panel of SNPs in the gene encoding COMT, they identified a haplotype that was associated with a 1.5-fold increase in odds of schizophrenia (95%CI = 1.4–1.6). The same haplotype has been shown to code for relatively lower expression of COMT mRNA in human brain tissue (Bray et al., 2003). Recently, we have identified 3 prevalent COMT haplotypes that accounted for 11% of variability in experimental pain perception in women, and which were predictive of TMD risk (Diatchenko et al., 2005). However, the potential for variations in this gene to mediate relationships among psychological characteristics, pain sensitivity, and TMD risk has not been examined. The overall goal of this study was to determine if psychological characteristics associated with pain sensitivity are predictive of TMD risk independently of any effects of COMT haplotype. We aimed to answer 4 related research questions:
We undertook a three-year, prospective cohort study of first-onset TMD among healthy, female volunteers aged 18–34 years at the time of recruitment. We aimed to follow 238 participants for up to 3 yrs, this being the number calculated to provide statistical power of 0.80 to detect risk ratios of at least 2.7, assuming a three-year cumulative incidence of 9%, which we estimated based on results reported by Von Korff et al.(1993). Prior to enrollment in the study, participants were screened and underwent a baseline physical examination of the head and neck, conducted according to the research diagnostic criteria (RDC) for TMD (Dworkin and LeResche, 1992). Individuals were excluded if they had a history of TMD, were diagnosed with TMD at the baseline examination, or if they reported a significant medical history including traumatic facial injuries or use of centrally acting medications (see APPENDIX Table 1). At baseline, peripheral blood samples were collected from enrolled participants, and they completed psychological questionnaires and psychophysical pain assessments. For up to 42 mos after their baseline assessment, participants were contacted every 3 mos by research staff, who administered a medical history update questionnaire. Any participants responding positively to key questions about TMD symptoms were re-examined according to the RDC protocol. Additionally, each year, all participants were invited to attend for RDC examination. New cases of TMD were defined according to RDC criteria (Dworkin and LeResche, 1992) for Group I Myofascial Pain and/or Group III Arthralgia, both of which include reported pain and pain on palpation. New cases were confirmed independently by two examiners (W.M., A.S.).
COMT Genotyping
The labels "sensitive" and "resistant" were based on our observed associations of these haplotypes with sensitivity to experimental pain, supported additionally by the results of in vitro and in vivo experiments (Diatchenko et al., 2005).
Psychological Measures
Experimental Pain Procedures Used to Index Pain Phenotype
Data Analysis
Ethical Conduct of Research
Two hundred and fifty-four females volunteered to take part in the study and completed baseline sensory assessments. Of these, 212 (83%) provided a blood sample and written consent for genotyping. Of those genotyped, 194 carried the 3 most abundant COMT haplotypes. Follow-up data about clinical TMD status were obtained from 171 (88%) of them. The percentage of people with follow-up data was higher for single women compared with those who were not single (P < 0.01, Table 1  ). However, rates of follow-up did not vary significantly among subgroups classified by any other characteristics listed in Table 1.
Fifteen new cases of TMD were diagnosed according to RDC criteria among the 171 participants who were followed for an average of 30 mos (range = 8–42 mos). The 15 new cases of TMD represented a cumulative incidence of 8.8% (95%CI = 5.0–14.2%), and the average annual incidence rate of TMD was 3.5 cases per 100 person-years (95%CI = 2.4–5.1 cases per 100 person-years). Among the 55 participants whose index of pain sensitivity had a value of 3.0 or more, the incidence rate was 5.9 cases per 100 person-years, compared with 2.4 cases per 100 person-years among the 116 people with lesser values, yielding a statistically significant IDR of 2.5 (95%CI = 1.2–5.2) (Fig. 1 ). This represented an elevation of risk that was similar in magnitude to our previously reported finding that the pain-sensitive COMT haplotype was predictive of a 2.3-fold increase in risk of TMD (Diatchenko et al., 2005; data included in Fig. 1 for comparison).
Bivariate correlations ranging from 0.15 to 0.25 in absolute value were observed between 10 psychological scales and the index of pain phenotype (Table 2 ). Although they represented only weak levels of correlation, 3 of them were statistically significant at the threshold of P < 0.0028: BSI depression, POMS confident-unsure, and Perceived Stress. However, the mean values of psychological variables generally did not differ significantly between the 2 COMT haplotype groups, suggesting little potential for confounding (Table 2). The one exception was BSI depression, which had a marginally higher mean value in the PSH group compared with the PRH group (P = 0.05) (Table 2).
When dichotomized at their approximate upper tertile, all 3 psychological variables identified in Table 2  were significant risk factors for TMD (Fig. 2). IDRs ranged from 2.6 for the Perceived Stress scale to 3.7 for the POMS confident-unsure subscale. However, the percentage of participants with PSH haplotypes did not differ significantly among psychological subgroups, as evidenced by prevalence ratios with 95%CIs that included unity, suggesting that they were not likely confounders of the haplotype-TMD association. As expected, the mean index of pain sensitivity differed significantly between each of the psychological subgroups (see APPENDIX, Table 4).
The BSI depression subscale was selected as the one psychological variable in which there was some potential for confounding due to COMT haplotype (Table 2 ). Multivariate analysis revealed independent effects on TMD risk of BSI depression (likelihood ratio statistic, 1 df, = 8.75, P = 0.004) and COMT haplotype (likelihood ratio statistic, 1 df, = 4.03, P = 0.046). In this multivariate model, which adjusted for COMT haplotype, the IDR for high BSI depression (subscale scores > 60) was 3.1 (95%CI = 1.5 6.4), compared with the reference group, which had low BSI depression scores ( 60). This IDR was not markedly different from the bivariate IDR of 3.2 for high BSI depression (Fig. 2).
In this prospective cohort study of otherwise-healthy females, pain phenotype and psychological characteristics associated with pain phenotype were both risk factors for first-onset TMD. The observed IDRs were comparable with risk ratios reported for other multifactorial conditions, such as schizophrenia (Shifman et al., 2002), and for TMD (Von Korff et al., 1993). Nonetheless, these represent only moderately strong predictors, highlighting our belief that no single gene or psychological characteristic is sufficient to explain variability associated with a complex condition such as TMD. The significance of these findings is strengthened by the prospective cohort study design, which overcomes a major limitation of previous case-control studies of TMD, in which it has been unclear whether putative risk factors such as pain sensitivity and depression existed in participants prior to the onset of TMD or arose as a consequence of it. Moreover, participants in this study were diagnosed independently by examiners using RDC guidelines. This provides confidence that the elevated risk of TMD is not simply an artifact of reporting bias among participants found to be at elevated risk. Our results suggest that effects of psychological characteristics on TMD risk cannot be attributed to variants of the gene encoding COMT. Mean psychological scores generally did not vary among COMT haplotype groups, and prevalence of the PSH haplotype did not differ significantly among groups dichotomized according to psychological subscale scores, suggesting little, if any, potential for confounding. Furthermore, results from the multivariate analysis of the one psychological subscale that showed some potential for confounding revealed little attenuation of the IDR for BSI depression after adjustment for COMT haplotype. The lack of attenuation demonstrated that the relationship between BSI depression and TMD risk could not be attributed to the COMT haplotype. The implication is that there are separate etiological mechanisms by which these psychological characteristics and genetic variants of the COMT gene influence the risk of a person developing TMD. When designing this study, we recognized that there would be a relatively small number of TMD cases diagnosed over 3 yrs. The observed cumulative incident of 8.8% is consistent with the rate of 7.7% observed during 3 yrs by Von Korff et al.(1993). Because there were few incident cases, we adopted a systematic analytic strategy to narrow the pool of candidate psychological characteristics by first screening for associations with pain phenotype, selecting only those variables that had statistically significant correlations. One consequence of this analytic strategy is that psychological characteristics not associated with experimental pain have been omitted as potential risk factors in this analysis. This is one reason that we found no association between TMD risk and somatization or anxiety—2 risk factors reported in some previous studies (Drangsholt and LeResche, 1999). The limited number of TMD cases also precluded any evaluation of interactions between genetic and these psychological characteristics. There are 2 principal clinical implications from these results. Most importantly, we interpret these findings as evidence of separate etiological pathways by which these psychological factors and COMT haplotypes increase risk of clinical pain. Second, the results highlight a need for a multifaceted approach to treatment and prevention of TMD that is based on an understanding of its etiology. For example, based on these findings, control of psychological characteristics and COMT activity may be effective strategies to prevent or treat TMD. Existing treatments for TMD, such as analgesics, dental-orthopedic devices, and surgery, often are decided on empirically, and sometimes despite evidence that some such treatments are no better than a placebo (Dao et al., 1994). Indeed, Svensson and Graven-Nielsen (2001) have asserted that "the pathophysiology and etiology of craniofacial muscle pain are not known in sufficient detail to allow causal treatment." The findings from this study, if confirmed in other populations, would provide a rationale for the development and evaluation of the efficacy of interventions for TMD that target psychological characteristics or that compensate for decreased COMT activity.
This study was supported by DE07509, NS045685, DE007333, AR/AI-44564, NIH Intramural Grants DE00366 and AA000301, and the Comprehensive Neuroscience Program Grant USUHS G192BR-C4.
A supplemental appendix to this article is published electronically only at http://www.dentalresearch.org. Received for publication August 14, 2005. Revision received October 1, 2006. Accepted for publication October 5, 2006.
Journal of Dental Research, Vol. 86, No. 11,
1120-1125 (2007) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
