Experimental pain and fatigue induced by excessive chewing

In total 38 persons were enrolled for inclusion examination and all were screened by one examinator (IV). Thirty-one healthy participants were found eligible and included in the study, 15 healthy men with a mean (SD) age of 27 (5.4) years and 16 healthy age-matched women aged 25 (4.3) years (Table 1), no one declined participation (Fig. 1). Twenty six participants were required according to a non-inferiority power calculation (https://​www.​sealedenvelope.​com/​power/​binary-noninferior/​) to achieve a significance level (α) of 0.05 and power (β) of 80% excluding a difference of more than 30% in pain intensity and duration in favor for the 60-min trial [11, 12].

The inclusion criteria were: a) age over 18 years; and b) good general health. The exclusion criteria were: 1) a diagnosis of myalgia, myofascial pain, arthralgia, headache attributed to TMD, degenerative joint disease, painful clicking or locking, all according to the Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) [4]; 2) additional palpatory tenderness of the masseter, temporal muscles or over the temporomandibular joint (TMJ); 3) clinically visible dental pathology or mobility, tooth wear grade 3 = exposure of pulp or secondary dentine according to the simplified scoring criteria for tooth wear index I [16], malocclusion, edentulous areas or dentures; 4) general chronic pain conditions, systemic inflammatory diseases (i.e. rheumatoid arthritis, fibromyalgia, etc.), neuropathic pain or neurological disease; 5) whiplash associated disorder; 6) use of any medication that might influence the response of pain i.e. analgesics during 24 h preceding the trial, use of cannabinoids, or any medication that might influence the neurological function; 7) self-reported bruxism and chewing gum for more than 30 min on a daily basis, since these activities may affect the chewing muscles’ resistance to fatigue [17]; 8) allergy to any of the contents in the chewing gum; 9) pregnancy; and 10) cognitive or physical disability that prevent participation.

Participants filled in questionnaires regarding psychosocial variables: anxiety (generalized anxiety disorder scale-7; GAD-7) [18], depression (the patient health questionnaire for depression-9; PHQ-9) [19], physical/somatic symptoms (the patient health questionnaire for physical symptoms-15; PHQ-15) [20], stress (perceived stress scale-10; PSS-10) [21] and pain catastrophizing (pain catastrophizing scale-13; PCS-13) [22]. Information regarding use of contraceptives and phase of the women’s menstrual cycle was obtained in order to take the hormonal variation into consideration [5]. All participants were clinically examined according to the standardized examination protocol of diagnostic criteria for temporomandibular disorders (DC/TMD-Axis I) prior to inclusion but also at the end of the chewing task as well as at the 1-h follow-up and the 2-h follow-up. Only one examiner (IV; trained in DC/TMD) performed all examinations, and she was blinded to the duration of the chewing trials. Counterbalancing was used to control for order effects. Therefore, participants were randomized, in blocks of four using a digital tool (www.​randomization.​com) and blinded to start either with a 40-min or a 60-min chewing trial and vice versa after a wash-out period of 1 week, by a researcher not participating in data collection (NCh). Hence, order effects would occur equally in both groups and balance each other out in the results. Based on pilot studies performed with chewing durations of 20, 40, 60 and 100 min (unpublished results) we chose to use duration of the chewing tasks limited to 40 versus 60 min. This choice was partly based on the outcome that there were no differences between the 60 and 100 min groups regarding any of the variables but there were unwished side-effects of intestinal load/discomfort after the 100 min trial. The 20-min group did not result in any induced pain or fatigue/exertion. Also, future clinical experiments should have a reasonable duration and consequently being performed the same day. Further, it is well-known that women are more susceptible to pain [5], a longer duration would result in a high number of drop-outs among them.

The participants chewed five new chewing gums divided in 5-min chewing bouts (ELMA® sugar free, Mastiha, Chios, Greece; 5 × 1,4 g = 7 g) [11, 12, 23, 24]. The participants were instructed to continuously chew without rest on their dominant habitual masticatory side, following their natural chewing pattern. The examiner IV monitored the chewing procedure during the entire task. In order to reduce the risk that a standardized rate could influence the results we chose to use the dominant habitual masticatory side and the participants’ natural chewing pattern. The gain of this method is that in daily life there are many individuals who are “fast-chewers” and if the chosen chewing rate in an experiment happens to be “slower” than those participants’ natural chewing rate then the results would be misleading since no subjective fatigue or pain would be detected [25].

The values of jaw subjective fatigue (Borg’s Rating of Perceived Exertion; Borg’s RPE) and pain intensity (Numeric Rating Scale; NRS) were monitored and assessed at baseline, every 10 min during, immediately after and every 10 min after the chewing task during the 1-h follow-up and every 20 min during the 2-h follow-up. Pain drawings were assessed at baseline, at the end of the chewing task and after 1 and 2 h respectively following the chewing tasks. Further, at baseline, every 20 min during, immediately after as well as every 20 min after the chewing task up till 2 h of follow-up, pressure pain thresholds (PPT) over the masseter and temporal muscles as well as the index finger (reference point), maximum voluntary bite force (MVBF) and maximum voluntary mouth opening capacity (max MOC) were assessed. One examiner (IV) performed all assessments and was blinded to the duration of the chewing trials. The experimental protocol and time points of measurement variables are illustrated in Fig. 2.

Subjective fatigue was assessed using Borg’s RPE [6‐20], where 6 is extremely easy effort and 20 is maximum effort [26].

Pain intensity and peak pain were assessed using a NRS (0–10) where the end-points were 0 = no pain and 10 = worst imaginable pain [27].

A lateral chart of the face for both the right and left sides separately as well as intra-orally was used for assessing the pain spread. The participants were asked to mark all the areas in which they sensed pain on the chart by drawing a ring around the painful space. The drawings were later scanned and the Adobe Photoshop CC software (version 19.1.3, Adobe Systems Incorporated, San Jose, CA, USA) was used to count the pixels within the marked total area in arbitrary units (au).

An electronic pressure algometer (Somedic Sales Hörby AB, Sweden) was used over the masseter and temporal muscles bilaterally to assess pressure pain threshold (PPT). The algometer is supplied with a soft rubber tip with a surface of 1 cm², which was applied perpendicular to the participants’ skin surface. The participants were asked to clench and relax in order to determine and mark the most prominent area of the masseter belly and the anterior temporal muscle which would be the site for the pressure application. The participants were also instructed to press a button immediately as the sensation of pressure turned into pain. The participants’ head was supported on the opposite side by the examiner’s hand. The pressure was increasing with a rate of 30 kPa/s [11, 28, 29]. The electronic pressure algometer was calibrated before each trial. PPT was assessed by one calibrated examiner (IV), and repeated twice over each muscle site at each assessment and the mean value was used for data analyses.

In order to assess maximum voluntary bite force in Kilogram (Kg), a bite force transducer (41.0 × 12.0 × 5.0 mm, length × width × height, Aalborg University, Aalborg, Denmark) was used. The bite force transducer was covered with 1 mm rubber in order to avoid any cross contamination and reduce the risk of tooth fracture and inserted between the first or second molars either on the right or left side depending on each participant’s dominant habitual masticatory side.

The maximum voluntary mouth opening capacity, inclusive the vertical overbite, was assessed according to DC/TMD-Axis I in millimeters.

The normality of all data was tested with the Shapiro-Wilk test. The data showed a non-normal distribution and a skewness to the right, all except age. Therefore, non-parametric tests were used to analyze data and all data except age are presented as median (interquartile range; IQR). The PPT, BF and MOC values were normalized and presented as the percentage change from baseline values. The data were analyzed with the SigmaStat software (version14.0; Systat Software Inc., San Jose, CA, USA) and for all tests, the level of significance was set at P <  0.05 for within groups comparisons and P <  0.005 for subjective fatigue and pain intensity, < 0.013 for pain area and <  0.006 for the rest of the variables for between groups comparisons after applying Bonferroni corrections. The data from the four participants that did not attend at the second session (all lost at the 60-min chewing trial) were analyzed for their first session, the 40-min chewing trial, and were handled as missing data for their second session, the 60-min chewing trial.

For baseline and between groups comparisons, Wilcoxon Signed Rank test was used to test differences between trials and Mann–Whitney Rank Sum test was used to test sex differences as well as testing differences between the sessions. For within groups comparisons, the nonparametric Friedman’s analysis of variance for repeated measures with Tukey post-hoc test for the associated multiple comparisons were used to test changes in all variables versus baseline. The factors included in the analyses were time (baseline, end of chewing task and follow-up time-points), trials (40-min chewing trial and 60-min chewing trial), sex (men and women) and sessions (session 1 and session 2).

The psychosocial characteristics of all participants are presented in Table 1. All psychosocial variables were within a normal range, except for the physical/somatic symptoms and stress in women that were of mild grade. The physical/somatic symptoms were also significantly higher in women than men. The baseline values of subjective fatigue, pain characteristics and functional measures are presented in Table 2. Women displayed lower baseline PPT, BF and MOC than men. There were no significant differences between those starting with the 40-min or 60-min chewing trial (Tables 1 and 2).

Excessive chewing induced a significant increase in subjective fatigue that lasted for 20 min after completed chewing, in both the 40 and 60 min trials when compared to baseline (Table 3 and Fig. 3a). There were no significant difference in subjective fatigue when the two chewing trials were compared (Table 4).

The 60-min task induced a significant increase in pain intensity (=peak pain) and pain area (Table 3), while the 40-min task did not. Nonetheless, this significant increase did not even last to the first 10-min follow-up (Fig. 3b). There were no significant changes in PPT over the masseter, temporal muscles or in the index finger (reference point) when compared to baseline values, neither in the 40-min nor in the 60-min trials (Table 3). The induced pain area was significantly bigger in the 60-min trial compared to the 40-min trial (Table 4). There were no significant differences in pain intensity or change of PPT over the masseter, temporal muscles or the index finger when the two trials were compared.

There were no significant changes regarding MVBF or maximum voluntary MOC after the 40-min chewing task or after the 60-min chewing task (Table 3), neither when the two chewing trials were compared (Table 4).

Differences between the sessions regarding subjective fatigue, pain characteristics and functional measures are presented in Table 5.

All the participants were healthy pain-free individuals at study start. Thus, no DC/TMD diagnoses can be made and the diagnoses presented below can therefore be considered either as acute DC/TMD- or DC/TMD-alike diagnoses.

Subjective fatigue increased significantly compared to baseline values in both men and women in 40-min and 60-min trials. No significant sex differences were found even if the fatigue-resistant slow type I fibers in masseter muscles have a significantly larger diameter in women than men, while for the type II fibers it is the other way around [55]. Type I motor units are recruited first [56] and muscle fibers with larger diameters produce faster action potentials [57]. The results seem to suggest that women recovered faster from the induced fatigue compared to men in the 60-min trial, which is probably be related to men expressing pain after the 60-min chewing task.

An interesting finding was that in men pain intensity increased significantly compared to baseline values in the 60-min trial. Still, no significant sex differences were found regarding the levels of pain intensity or pain area in both 40-min and 60-min trials. This is in line with results from a previous study [42] that also showed no sex differences in pain intensity and area. However, these results are contradictory to those from a previous study where women reported a higher level of pain intensity as well as a larger pain area [58]. Perhaps the sensation of subjective fatigue exceeded the sensation of pain and thereby affecting the men’s subjective assessment of pain. No sex differences could be shown regarding the changes in PPTs in our study. This is in line with previous studies where PPTs remained unchanged in women [12] and in men [32] after excessive chewing or clenching. However, there is an inconsistency in results from previous studies since there were other studies showing that women had lower PPT values than men [58] or a significant decrease in PPT until 24 h after a 100 min chewing task in men [11]. An explanation to the discrepancies could be the different methodologies. Moreover, sex differences are more obvious in longer-lasting pain conditions [59].

The clinical examination was performed by one examiner only who was blinded to the duration of the trial and used the standardized and robust protocol DC/TMD [4]. The study was divided into two sessions with a wash-out period of 1 week in order to avoid any contamination of the data due to DOMS [8, 9, 60]. The included men and women were age matched in order to analyze any possible sex differences [58]. There were further no differences in any of the bio-psycho-social variables between the participants who started with 40-min or 60-min chewing, which confirmed the counterbalancing and enhanced the homogeneity of the population recruited.

One possible limitation could be the fact that we did not have enough material to account for the menstrual cycle. The phases of menstrual cycle were asked for but not taken into consideration since as all participating women were in different menstrual phases. Hence this study could be seen as representative for the population and it had been showed that intra-individual variability in the pain response is greater than the influence of estrogen [61]. The self-reported bruxism could be considered an information bias since nightly bruxism could still occur without the individual’s knowledge. The examiner and the participants could count the amount of chewing bouts to find out which was the longer duration trial, and thus not being completely blinded. During-chewing-data were assessed, however, in order to avoid any confusion between during task and follow-up time-points we chose not to present those data. Women seemed to recover faster form the induced fatigue after the 60-min trial compared to the 40-min trial which might appear puzzling. A possible selection bias that might occur due to women dropping-out from the second session at which they were randomized to the 60-min chewing trial could explain the finding and would be considered a drawback.