Temporomandibular disorders (TMD) are quite common among the general population, with a lifetime prevalence of up to 93% in an epidemio- logical study (1). These disorders include com- plaints of the temporomandibular system, consisting of the temporomandibular joint (TMJ) and the associated neuromuscular system (2). A national survey among Dutch adults showed that 21.5% of the adult population had temporomandib- ular dysfunction but only 15% of those sought treatment (3). Most patients seek treatment because of TMD pain (4), in the temporomandibular region or involving the eyes, face, shoulder, neck, back, and head (5, 6). Patients with TMD have a decreased quality of life because of orofacial pain, particularly for patients with severe TMD (7). The etiology of TMD has been regarded as multifactorial. Dworkin
and LeResche (8) designed a two-axis diagnostic scheme to evaluate the patient’s condition. The psychological variables were assessed with Axis II, emphasizing the relevant factors of TMD. Gracely (9) reported that individuals can have different levels of pain perception, which can be influenced by emotional factors. Hotopf et al. (10) have noted the psychiatric disorder may promote 40% cases of multiple symptoms including arthritis, rheuma- tism, and headache. Magni et al. (11) found in a prospective study that the relationship between depressive systems and chronic musculoskeletal pain may operate in both directions.
Depressive symptoms are significantly related to the severity of pain in the TMD patients (12).
Moreover, TMD pain and depression are often co-existent (13–15). Macfarlane et al. (15) found in a
The risk of temporomandibular disorder in patients with
depression: a population-based cohort study
Liao C-H, Chang C-S, Chang S-N, Lane H-Y, Lyu S-Y, Morisky DE, Sung F-C.
The risk of temporomandibular disorder in patients with depression: a population-based cohort study. Community Dent Oral Epidemiol 2011; 39:
525–531. 2011 John Wiley & Sons A ⁄ S
Abstract – Objectives: This study used a population-based retrospective cohort design to examine whether depression is a risk factor of temporomandibular disorder (TMD). Methods: From a universal insurance database, we identified 7587 patients who are newly diagnosed individuals with depression in 2000 and 2001. A total of 30 197 comparison subjects were randomly selected from a nondepression cohort. Both groups were followed until the end of 2008 to measure the incidence of TMD. Results: The incidence of TMD was 2.65 times higher in the depression cohort than in the nondepression cohort (6.16 versus 2.32 per 1000 person-years). The hazard ratio (HR) measured by multivariate Cox’s proportional hazard regression analysis of TMD for the depression cohort was 2.21 (95% confidence interval (CI) 1.83–2.66), after controlling for socio- demographic factors and other psychiatric comorbidities. Women had higher risk to develop TMD than men (HR 1.61, 95% CI 1.36–1.92 for women without depression; HR 3.54, 95% CI 2.81–4.45 for women with depression). Conclusions:
This study demonstrates that patients with depression are at an elevated risk of developing TMD.
Chun-Hui Liao
1, Chen-Shu Chang
2, Shih-Ni Chang
3,4,5, Hsien-Yuan Lane
1,6, Shu-Yu Lyu
7, Donald E. Morisky
8and Fung-Chang Sung
4,91
Department of Psychiatry, China Medical University and Hospital, Taichung, Taiwan,
2
Department of Neurology, Changhua Christian Hospital, Changhua City, Taiwan,
3
The Ph.D. Program for Cancer Biology and Drug Discovery, China Medical University, Taichung, Taiwan,
4Management Office for Health Data, China Medical University, Taichung, Taiwan,
5Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan,
6
Institute of Clinical Medical Science, China Medical University and Hospital, Taichung, Taiwan,
7School of Public Health, Taipei Medical University, Taipei, Taiwan,
8
Department of Community Health Sciences, UCLA School of Public Health, Los Angeles, CA, USA,
9Department of Public Health, China Medical University and Hospital, Taichung, Taiwan
Key words: depression; hazard ratio;
population-based study; retrospective cohort design; temporomandibular disorder Fung-Chang Sung, China Medical University and Hospital College of Public Health, 91 Hsueh-Shih Road, Taichung 404, Taiwan Tel.: +886 4 2206 2295
Fax: +886 4 2201 9901 e-mail: fcsung1008@yahoo.com Submitted 31 August 2010;
accepted 18 April 2011
case–control study that patients with pain dysfunc- tion syndrome had high levels of psychological distress. Depression has now become a global burden disorder and the fourth leading cause of disability worldwide (16). Major depressive disor- der presents in 5–10% of patients seeking primary care (17). The prevalence of depression may well be higher among the general population because some people may have depressive disorders which do not fully meet the diagnostic criteria for major depressive disorder of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition.
Among studies on the relationship between depressive systems and pain, there is a convergent association between depression and TMD pain, especially in the chronic pain group (2). Several studies on the relationship between depression and TMD have evaluated the mood condition among TMD patients (13–18). However, those studies were unable to answer whether depression is a source or consequence of TMD pain because of case–control and cross-sectional designs.
Slade et al. have recently conducted a prospective cohort study of 238 healthy female volunteers aged 18–34 years investigating the psychological influ- ence on the risk of TMD. They found that depres- sion, perceived stress, and mood are associated with pain sensitivity with a two- to threefold increase in the risk of TMD (P < 0.05) (19). But, this study was limited to a small sample size with only one gender and a short follow-up period. We therefore designed a population-based cohort study with higher statistical power to detect the development of TMD among the depressive patients. Not all patients with TMD complaints visit the dentists. We hypothesized that patients with depressive disorder would have a higher risk of TMD complaints, leading to a greater likelihood to seek dental services than the general population. To gain a better under- standing of the relationship between depression and TMD, we conducted a population-based retrospec- tive cohort study using claims data from the universal insurance program. The incidences of dentist-diagnosed TMD were compared between patients with depression and without depression.
Materials and methods Data resources
This study used the reimbursement claims data of The National Health Insurance program of Taiwan
that reformed in March 1995 from 13 insurance systems. The insurance program has covered more than 96% of the 23 million population and con- tracted with more than 90% of hospitals and clinics in Taiwan since 1996 (20, 21). The Department of Health National Health Research Institute (NHRI) managed all medical claims data reported from the contracted health care facilities. With approval from NHRI, we were able to use a representative sub-datasets of 1 million insured persons ran- domly selected from all beneficiaries enrolled in the insurance program (21). This dataset consisted of the registry of medical facilities, details of inpatient orders, ambulatory care, dental services, and prescriptions linked with scrambled patient identification. Because all patient identifications were surrogated, this study was conducted with patients privacy secured and with a waiver from the institutional review board.
Study sample
We used the coding of the International Classifica- tion of Disease Diagnoses, Ninth Revision of Clinical Modification (ICD-9-CM), to identify 7587 patients with depression (ICD-9-CM 296.2, 296.3, 300.4 and 311) newly diagnosed in 2000 and 2001 as the study cohort. To ensure the validity of the diagnosis, only new patients with at least three visits for depression care during the follow-up period after the index date were eligible for inclusion. For each case with depression identified, we used simple random sampling methods to select four persons without depression in the same period for the comparison cohort (N = 30 348). We excluded 102 subjects from the depressive cohort and 151 subjects from the comparison cohort. They were excluded because of a history of TMD diagnosis by the baseline index date (defined as the date the subject identified and selected) or missing information on age or sex. Our final sample includes 7485 subjects in the depression cohort and 30 197 subjects in the nondepression cohort.
Socio-demographic variables and comorbidities
The socio-demographic variables, including sex, age, occupation, employment category, residential area, and monthly income, were available. The age of each study subject was measured by the differ- ence between the index date and the date of birth.
Using the National Statistics of Regional Standard
Classification (22), we grouped all study subjects
into four geographic areas (North, Central, South,
and East and off Islands) and three urbanization levels (low, medium, and high).
We considered anxiety state (ICD-9-CM 300.00), panic disorder (ICD-9-CM 300.01), generalized anxiety disorder (ICD-9-CM 300.02), obsessive- compulsive disorders (ICD-9-CM 300.03), and psy- chiatric diseases (ICD-9-CM 290-319, except the main effect in this study – depression) as other psychiatric comorbidities.
Study end point
We linked study subjects to the inpatient and outpatient claims data of dental clinics to identify the newly diagnosed cases of TMD (ICD-9-CM 524.6) as the outcome of the study, using the scrambled patient identification number. We cal- culated person-years for each study subject until TMD was diagnosed, or until December 31, 2008, for those uncensored, or the censoring date because of death, emigration, termination of insurance, or loss to follow-up.
Statistical analysis
We compared the distributions of categorical socio- demographic variables and comorbidities between depression patients and nondepression patients using the Chi-square test. We also calculated the incidence density with person-years for these variables in the study cohort and comparison cohort. The rate ratio of TMD was calculated by each variable.
Cox’s proportional hazard regression analysis was used to assess the risk of TMD associated with depression, adjusting for variables that were sig- nificantly related to depression from the prior Chi- square analyses. Hazard ratio (HR) and 95%
confidence interval (CI) were calculated in the model. The sex and age stratification analyses for the risk of TMD in association with depression were also examined using Cox’s proportional hazard regression analysis.
All analyses were performed with sas statistical software (version 9.1 for Windows; SAS Institute, Inc., Cary, NC, USA). Significance level was set to 0.05.
Results
Table 1 compares socio-demographic characteris- tics between the depression cohort and the nonde- pression comparison cohort. There were more women in the depression cohort than in the
comparison cohort (60.6% versus 48.1%). The depression cohort was also older, less white collar employment and had middle income.
Table 2 presents the incidence and crude hazard ratios of TMD by socio-demographic status. The overall incidence rate of TMD in the depression cohort was 2.65 times higher than that in the comparison cohort (6.16 versus 2.32 per 1000 per- son-years). The crude hazard ratios measured by categorized socio-demographic status ranged from 2.35 to 3.64, with the depression cohort of those more than 60 years of age having the highest hazard ratio. Older men in the nondepression cohort had the lowest risk of having TMD.
The multivariate Cox proportional regression analysis showed that the risk of TMD was signif- icantly greater in the depression cohort than in the Table 1. Comparisons in demographic characteristics between patients with and without depression at base- line in 2000–2001
Variables
Depression*
No
aN = 30 197
Yes N = 7485
n (%) n (%)
Sex
Female 14 529 (48.1) 4536 (60.6)
Male 15 668 (51.9) 2949 (39.4)
Age, years
<20 9240 (30.5) 363 (4.9)
20–39 10 558 (35.0) 2307 (30.8)
40–59 6995 (23.2) 2672 (35.7)
‡60 3404 (11.3) 2143 (28.6)
Occupation
White collar 16 977 (56.2) 3440 (46.0) Blue collar 9412 (31.2) 2675 (35.7)
Others
b3808 (12.6) 1370 (18.3)
Urbanization
aLow 3627 (12.1) 1007 (13.5)
Moderate 6337 (21.1) 1458 (19.6)
High 20 092 (66.9) 4978 (66.9)
Region
North 14 215 (47.0) 3148 (42.1)
Central 6162 (20.4) 1534 (20.5)
South 7535 (25.0) 2104 (28.1)
East and Island 2284 (7.6) 699 (9.3) Income
<15 000 15 164 (50.2) 3138 (41.9) 15 000–29 999 11 257 (37.3) 3505 (46.8)
‡30 000 3776 (12.5) 842 (11.3)
Depression: ICD-9 codes: 296.2, 296.3, 300.4, and 311.
a
Urbanization: low = 1st and 2nd lowest quartile of population density, moderate = 3rd quartile of popula- tion density, high = 4th highest quartile of population density.
b
Unemployed, retired and low income.
*Chi-square test, all P values are <0.001.
nondepression cohort (HR 2.21, 95% CI 1.83–2.66) after controlling for covariates (Table 3).
Table 4 shows that women were at greater HRs to develop TMD than men. Compared with men without depression, women without depression had a HR of 1.61 (95% CI 1.36–1.92) and women with depression had the HR increased to 3.54 (95%
CI 2.81–4.45). Compared with the nondepression cohort 360 years of age, the HRs of TMD increased in patients with depression, and in
particular, the highest risk was noted in depression patients 360 years of age (HR 3.22, 95% CI 2.20–
4.73).
Discussion
Studies investigating whether the risk of TMD is higher in patients with depression using cohort designs are limited. Our study aimed at exploring
Table 3. Hazard ratio of temporomandibular disorder in association with depression in Cox’s proportional hazard models
Variables
Model 1 Model 2 Model 3
HR (95% CI) HR (95% CI) HR (95% CI)
Depression
No 1.00 (reference) 1.00 (reference) 1.00 (reference)
Yes 2.64 (2.30–3.03)* 2.42 (2.09–2.81)* 2.21 (1.83–2.66)*
Model 1: unadjusted.
Model 2: adjusted for age, sex.
Model 3: adjusted for age, sex, area, occupation, urbanization, income, and other psychiatric comorbidity.
*P < 0.0001.
Table 2. Comparisons of incidence of temporomandibular disorder between cohorts with and without depression by socio-demographic factor
Variables
Depression
Crude HR (95% CI)
No Yes
Cases Person-years Rate
aCases Person-years Rate
aAll 533 229 485 2.32 338 54 872 6.16 2.64 (2.30–3.03)
Sex
Female 319 110 865 2.88 241 33 785 7.13 2.49 (2.10–2.95)
Male 214 118 619 1.80 97 21 087 4.60 2.49 (1.95–3.18)
Age, years
<20 149 72 808 2.05 15 2787 5.38 2.78 (1.63–4.73)
20–39 207 79 699 2.60 109 17 558 6.21 2.36 (1.86–2.99)
40–59 137 53 806 2.55 123 20 276 6.07 2.35 (1.84–3.01)
‡60 40 23 172 1.73 91 14 252 6.39 3.64 (2.50–5.30)
Occupation
White collar 313 129 644 2.41 172 25 691 6.69 2.70 (2.23–3.26)
Blue collar 163 71 273 2.29 105 19 628 5.35 2.42 (1.89–3.10)
Others 57 28 569 2.00 61 9553 6.39 3.15 (2.19–4.53)
Urbanization
Low 57 27 151 2.10 44 7138 6.16 2.93 (1.97–4.36)
Moderate 109 47 877 2.28 74 10 664 6.94 3.01 (2.23–4.07)
High 367 153 372 2.39 219 36 748 5.96 2.48 (2.09–2.94)
Region
North 244 107 867 2.26 141 23 217 6.07 2.61 (2.11–3.23)
Central 110 46 962 2.34 65 11 327 5.74 2.39 (1.75–3.27)
South 145 57 316 2.53 107 15 253 7.02 2.84 (2.21–3.66)
East and Island 34 17 332 1.96 25 5075 4.93 2.57 (1.54–4.32)
Income
<15 000 260 116 460 2.23 135 22 391 6.03 2.70 (2.19–3.34)
15 000–29 999 196 83 606 2.37 162 26 066 6.21 2.65 (2.15–3.27)
‡30 000 77 29 419 2.62 41 6415 6.39 2.35 (1.59–3.47)
a
Per 1000 person-years.
whether depression is a risk factor related to subsequent TMD problems. We measured the incidence of dentist-diagnosed TMD in depressive patients compared with that of a nondepression cohort, using a population-based retrospective cohort study design. This approach overcomes the major limitation of cross-sectional and case–
control study designs by providing incidence information.
Temporomandibular disorder disorders have a high degree of comorbidity with depression (13, 14, 23). Previous studies on TMD and depression have been unclear as to whether depression occurred prior to the onset of TMD or as a consequence of it.
Slade et al. (19) found that depression was one of the predicted risk factors of the first onset of TMD among healthy women with a small sample size.
These studies collected information on depression based on self-reported questionnaires, which is different from clinically verified depressive disor- der. Our population-based study on the association between depression and subsequent TMD found that there is a 2.21- to 2.64-fold higher risk of a diagnosis of TMD among patients with a physi- cian-diagnosed depressive disorder, compared with the control group within an 8-year follow-up period, after adjusting for demographic character- istics and comorbid anxiety disorders.
Temporomandibular disorder are a heteroge- neous group of disorders affecting TMJ, the mas- ticatory muscles, or both, and might present with joint sounds or severe dysfunction. The most common symptom was pain, and most patients sought help because of it. TMD has also been a chronic pain condition in many cases (24). Not only is depression prevalent among patients with
chronic TMD-related pain conditions, but patients with TMD and comorbid psychological factors had a poor response to dental treatment alone (25). This could be explained by depression possibly increas- ing pain-perception thresholds (26, 27) and affect- ing the expression of TMD signs and symptoms (28). Therefore, depressive patients might have more TMD problems, which lead them to seek dental services.
In our study, the incidence of TMD among the depression group was 4.5% in the 8-year follow-up, which is higher than the incidence in a previous study (3.1%) for a Dutch adult population (3). This reflects a higher risk of TMD problems among the depressive population. Our definition of TMD was based on the information from patients who had visited a dental clinics rather than from case–
control study or general cross-sectional survey among the general population. Therefore, in our study, the difference in TMD incidences between the depressive patients and the general population without depression is more valid. However, our diagnosis of depression included both minor and major depressive disorders (ICD 296.2, ICD 296.3, ICD 300.4, and ICD 311) and may have identified subjects with a broader spectrum of depressive disorders. It still is likely that depression is under- estimated in the general population, because not all patients seek help from physicians when de- pressed.
Our study was compatible with the previous reports that found that women had a higher rate of TMD than men (29–31). Elderly depressive patients are also found to have a higher risk of TMD, which lead them to seek dental services. However, the nondepressive elderly had the least risk of TMD.
Further research to evaluate TMD among geriatric patients is needed.
There are some limitations to interpret the results of our study. First of all, the diagnoses of depres- sion, TMD, and comobidity relied on claims data, so there may be missing information made under a standardized diagnostic process. Obtaining this kind of information for a large population-based cohort study would be extremely difficult. But our strength in this study was that working from a clinical diagnosis made it possible to avoid the limitations of self-reported questionnaires. To increase the diagnostic validity, all cases were diagnosed with depressive disorder at least three times, which provided a reliable cohort assessment.
Second, some studies found that a myofascial type of TMD have a higher comorbidity of depression Table 4. Sex- and age-specific hazard ratio of temporo-
mandibular disorder associated with depression mea- sured with multivariable Cox method
Variables
Depression No
aHR (95% CI)
Yes
aHR (95% CI) Sex*
Female 1.61 (1.36–1.92)*** 3.54 (2.81–4.45)***
Male 1.00 (reference) 2.23 (1.70–2.94)***
Age, years
a<20 1.25 (0.87–1.81) 2.70 (1.48–4.94)**
20–39 1.55 (1.09–2.19)* 3.05 (2.10–4.45)***
40–59 1.49 (1.04–2.13)* 2.98 (2.05–4.33)***
‡60 1.00 (reference) 3.22 (2.20–4.73)***
a