| Sign In to gain access to subscriptions and/or personal tools. |
Basal Circadian Cortisol Secretion in Women with Temporomandibular Disorders
1 Department of Psychological Medicine, University of Wales College of Medicine, Heath Park, Cardiff CF14 4XN, UK; Correspondence: *corresponding author, akorszun{at}umich.edu
Muscular temporomandibular disorder (TMD) is a common stress-related condition showing marked comorbidity with depression and fibromyalgia (FM), both of which are associated with dysregulation of cortisol secretion. We measured cortisol levels in 15 women with well-defined TMD and 15 matched controls by sampling blood at 10-minute intervals over 24 hours in a controlled environment. TMD patients showed markedly increased daytime cortisol levels 30% to 50% higher than those of controls (p = 0.0032) and a one-hour phase delay in the timing of maximum cortisol levels (p = 0.048). Increased activation of the stress hormone axis by conscious pain perception is a likely explanation, but the magnitude of the increase could indicate that pain in the facial region acts as a greater stimulus than pain elsewhere in the body.
Key Words: cortisol • temporomandibular disorders • facial pain
Chronic facial pain affects up to 6% of the adult population (Lipton et al., 1993) and is most commonly due to temporomandibular disorder (TMD), which can involve the temporomandibular joint and/or the masticatory muscles. However, in the majority of cases, the only clinical finding is tenderness on palpation of the muscles of mastication (Marbach and Lipton, 1982; Stohler, 1995; Korszun et al., 1996). Although the underlying cause of TMD remains poorly understood, it is widely recognized to be multifactorial, involving physiological, behavioral, and environmental factors. There is a marked comorbidity of TMD with depression (Feinman, 1983; Gallagher et al., 1991; Korszun et al., 1996), and both the onset and exacerbation of symptoms are associated with the occurrence of environmental stressors. Thus, TMD is regarded as one of a group of stress-related disorders that includes fibromyalgia (FM) and chronic fatigue syndrome (CFS) (Korszun et al., 1998). Dysregulation of stress response hormones is well-documented in depression (Holsboer, 1999) and has also been demonstrated in FM and CFS (Demitrack, 1998), although the pattern of abnormalities differs from that found in depression. We hypothesized that patients with TMD may also exhibit abnormalities of stress response hormones, and if so, this would provide further evidence of a biologically defined group of stress-related somatic syndromes. The main neuroendocrine component of the stress response is the hypothalamic-pituitary-adrenal (HPA) axis. Cortisol is the principal circulating glucocorticoid, and its release by the adrenal gland is controlled by adrenocorticotropin (ACTH) produced by corticotrophs in the anterior pituitary gland. The pituitary is in turn under the control of hypothalamic corticotropin-releasing hormone (CRH). HPA axis activation is subject to negative feedback regulation mediated via glucocorticoid receptors within both the brain and the anterior pituitary. Variations in plasma levels of ACTH and cortisol occur as a result of a pronounced circadian rhythm, as well as stress-induced HPA axis activation. The circadian, or near-24-hour, rhythm in plasma cortisol levels is very robust, and peak concentrations occur in the early morning hours, falling progressively to a nadir around the early to midpoint of sleep. Stress-induced secretion is superimposed on the basal circadian rhythm (Chrousos and Gold, 1992). The aim of this study was to test the hypothesis that patients with muscular TMD would show alterations in basal cortisol secretion. We measured cortisol levels over a 24-hour period in a well-defined group of TMD patients and individually matched controls, using an experimental paradigm involving frequent blood sampling through an intravenous catheter in a controlled environment.
Subjects Fifteen pre-menopausal women with muscular TMD only (Dworkin and LeResche, 1992) were recruited by advertisement. All participants gave informed consent, and the study was approved by the University of Michigan Institutional Review Board. Nine subjects were on oral contraceptives (OC) while six had normal menstrual cycles. Fifteen control subjects were individually age-matched (± 2 yrs) and, for those subjects taking OC, were matched to be on the same type of OC. All patients and controls were Caucasian non-smokers with a body mass index < 30 and were screened to exclude any recent chronobiological disruptions, such as shift work or travel across time zones. All participants were evaluated by structured psychiatric interview (SCID-IIIR or IV), and no comorbid Axis I disorders existed except for three subjects with major depression. All subjects were medication-free except OC as above. All psychoactive medications were discontinued for at least 2 wks prior to study (fluoxetine 2 mos prior to study). Those not taking oral contraceptives were all studied in the follicular phase of the menstrual cycle (within 10 days after the onset of menstruation).
Procedures Subjects also completed the Beck Depression Inventory (BDI) and McGill Pain Questionnaire (MPQ). Subjects were admitted to the GCRC on the evening prior to 24-hour sampling so that they could equilibrate to the unit. Meals were served at regular times (0730, 1200, and 1730 hrs, respectively), and lights were turned off between 2200 and 0600. During the procedure, subjects were required to rest quietly in a bed or chair, and television and radio use was prohibited. At 0700 hr the following morning, a large-bore intravenous catheter was placed in an antecubital or other large arm vein and attached by a connection to a double-stopcock assembly attached to an intravenous (iv) line for withdrawal of samples. Intravenous fluids (0.45% saline) were infused to keep the catheter open between sample withdrawals, and heparin (1000 u per liter) was added to iv fluids if blood flow diminished during sample withdrawal. Samples of 2.8 mL of venous blood were collected every 10 min over 24 hrs, yielding a total of 406 mL. Blood was collected in pre-chilled, sequentially labeled polypropylene tubes containing EDTA to prevent clotting and inhibit proteolytic activity. Plasma was separated by centrifugation within 2 hrs of collection, then frozen and stored at -70°C until assayed.
Hormone Assays
Statistical Analysis
To test if there were shorter periodicities in the cortisol series, we performed an ultradian analysis on the control group by regressing standardized hormone concentration values on sine and cosine terms representing the following periods: 24-hour, 12-hour, eight-hour, six-hour, three-hour, two-hour, one-hour, and To test whether there was an underlying circadian pattern present in the data and whether this pattern differed between the patients and controls, we fitted regression models to each individual's data, using the significant circadian (24- and 12-hour) terms (cosinor analysis). A paired t test was performed on the maximum time, the logged (base-e) difference between the maximum and minimum predicted values, and the logged (base-e) ratio of the maximum and minimum predicted values, so that we could determine if these measures differed by patient/control status after adjustment for oral contraceptive status.
Effects of Oral Contraceptives on Mean Cortisol Levels Cortisol levels were significantly higher in patients and controls on OC compared with those in patients and controls not on OC throughout the 24-hour period (p = 0.0012; Fig. A  ; Table 1).
Mean Cortisol Levels There was a significant effect of time in cortisol concentrations in both patients and controls (p = 0.0001, Table 1  ); mean cortisol peaked at the 6:00-9:50 time period and then declined (Fig. B). However, this time effect differed for patients and controls (p = 0.0032, Table 1). Cortisol levels were significantly higher in TMD patients for the three time periods: 10:00-13:50, 14:00-17:50, and 18:00-21:50 (Table 2; Bonferroni adjusted for multiple comparisons). On average, both patients and controls peaked during the time period 6:00-9:50, but patients remained at a higher level for a longer time than the controls. Though subjects using OC had higher mean cortisol levels, the time effect and patient-control differences remained the same regardless of OC use (Table 1).
Circadian and Ultradian Analysis Ultradian analysis revealed significant periods of 24 hrs, 12 hrs, which describe the circadian pattern, and 2 hrs. Thus, there is a shorter periodicity in cortisol levels in addition to the circadian pattern. From the cosinor analysis, the maximum time of hormone secretion (acrophase) in controls was 7:28 ± 18 min, while it was 8:32 ± 19 min for patients, a significant difference (p = 0.048, Table 3 ), after adjustment for contraceptive use. There were no significant differences between patients and controls for amplitude of cortisol secretion or in the ratio of maximum to minimum predicted hormone concentrations. Oral contraceptives demonstrated no effect on the maximum time but significantly increased the overall amplitude of the circadian rhythm of cortisol (p = 0.0001, Table 3).
Effect of Depression on Cortisol Levels Only three subjects in the present study fulfilled the criteria for major depression. The pattern of HPA axis activation did not differ between subjects with and those without major depression (Fig. C compared with Fig. B).
Sleep and Pain in TMD
TMD patients showed markedly increased daytime cortisol levels compared with controls as well as a one-hour delay in the timing of maximum cortisol levels. However, circadian amplitude and ratio did not differ between TMD patients and controls. There was a clear effect of OC on cortisol levels, which was manifested as a uniform increase in cortisol levels over the entire 24-hour period and which occurred in both subjects and controls. This can be explained by the known action of OC on cortisol binding globulin levels and emphasizes the need for inclusion of a carefully matched control group. The elevated cortisol levels in TMD patients most likely indicate central activation of the HPA axis, an explanation also hypothesized for patients with more generalized chronic pain syndromes. Central HPA axis hyperactivity could represent a response to pain, and this explanation is supported by the observation that the time of increased cortisol secretion corresponded to the time when individuals were awake and presumably aware of their pain. There were no differences between patients and control subjects during sleep. However, it is noteworthy that patients in this study had very minimal levels of pain, and although McGill pain scores were higher than in controls, the scores were not markedly raised compared with other chronic pain states. Subjects in this study did not show significant functional impairment and were able to pursue their normal activities. Furthermore, most subjects were recruited by advertising rather than by presenting as patients, indicating a potentially less morbid form of TMD. The marked increase in morning cortisol levels is in contrast to the findings in a recent study of circadian rhythms in FM patients with generalized muscle pain, where there was no circadian pattern in pain perception and cortisol levels did not differ from those of control subjects (Klerman et al., 2001). It remains possible that pain of the facial region represents a greater stimulus to HPA axis activation than pain elsewhere in the body. For example, one study of women with atypical facial pain demonstrated differences in cerebral responses to pain compared with responses in normal controls, with an increased blood flow in the anterior cingulate cortex and decreased blood flow in the prefrontal cortex (Derbyshire et al., 1994). Studies directly comparing TMD with disorders characterized by more generalized pain will be necessary for this possibility to be addressed. Activation of the HPA axis in acute stress is usually associated with analgesia, and CRH can affect pain processing both centrally and peripherally (Lariviere and Melzack, 2000). Furthermore, CRH has been shown to produce analgesia in post-operative dental pain, particularly in the affective component of pain (Hargreaves et al., 1987). In contrast to acute stress, it has been suggested that chronic exposure to stress may result in hyperalgesia (Lariviere and Melzack, 2000). It is possible that high levels of cortisol in TMD represent a physiologic response to chronic stress, with pain as a potential stressor, associated with chronically increased CRH or other central mediators of the HPA axis. Increased activation of the central components of the stress axis, perhaps with resulting hyperalgesia, is compatible with our data of increased cortisol secretion in TMD. The fact that high cortisol levels occur during the daytime only may reflect a decreased resiliency of the HPA axis in patients with TMD to the stimulus of consciously perceived pain. Another possible reason for the higher daytime levels of cortisol and later maximum time could be disruption of circadian rhythms, such as the sleep-wake cycle. However, activity monitoring before and during the study and patient log information showed no evidence of any phase shift. Patients with TMD usually complain of disturbed sleep, but, according to patients' reports, as well as their actigraphy, the present group of subjects did not show sleep disruption. This is in accordance with actigraphy data from FM patients who, although complaining of disturbed and unrefreshing sleep, showed a greater than 90% sleep efficiency (Korszun et al., 2002). While TMD is frequently comorbid with depression, we do not believe that depression can account for the increased cortisol levels seen in this study. Only three of the 15 subjects met the criteria for major depression, and patients without depression had significant cortisol elevation. Furthermore, the pattern of cortisol secretion in TMD patients is different from that seen in depression, which has generally been reported to be either uniform activation or greater activation in the late evening or nadir of the circadian rhythm. Analysis of our data supports the hypothesis that alterations of stress response hormones occur in several somatic syndromes associated with symptoms of chronic pain and indicates that pain in the facial region acts as a greater stimulus than pain elsewhere in the body.
This study is supported by NIDR RO1 DE11972-01 and by University of Michigan General Clinical Research Center NIH MO1-RR00042. Received for publication July 3, 2001. Revision received January 28, 2002. Accepted for publication February 6, 2002.
Journal of Dental Research, Vol. 81, No. 4,
279-283 (2002) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
-hour (
