Received: 1 March 2019 | Revised: 9 May 2019 | Accepted: 13 May 2019

DOI: 10.1111/joor.12823

ORIGINAL ARTICLE

Effects of 8 weeks of masticatory muscles focused

endurance exercises on women with oro‐facial pain and
temporomandibular disorders: A placebo randomised
controlled trial

Michelle Almeida Barbosa1,2 | Ariany Klein Tahara1,2 | Isabella Christina Ferreira1 |

Leonardo Intelangelo3 | Alexandre Carvalho Barbosa1,2,4

1
Musculoskeletal Research Group—NIME,
Department of Physical Therapy, Federal Abstract
University of Juiz de Fora, Governador Background: Exercises are used to treat temporomandibular disorders (TMD), but
Valadares, Brazil
2 they are often assessed with other therapies. Local endurance exercises may alter
Master Degree Program in Rehabilitation
and Physical Performance, Federal the resistance to fatigue and pain.
University of Juiz de Fora, Juiz de Fora,
Objective: To assess the effects of an 8‐week protocol of local endurance exercises
Brazil
3
Department of Physical
of masticatory muscles on muscle excitation, force response, perceived pain and over
Therapy, Universidad del Gran Rosario, muscle efficiency.
Rosario, Argentina
4
Design: Randomised controlled trial.
Master Degree Program in Applied Health
Sciences, Federal University of Juiz de Fora, Setting: Ambulatory care.
Governador Valadares, Brazil Subjects: In a placebo randomised controlled trial, 46 women with TMD and oro‐facial
Correspondence pain were randomised into intervention group and placebo group. The intervention
Alexandre Carvalho Barbosa, group received a protocol of biting endurance exercises, controlled by biofeedback.
Musculoskeletal Research Group—
NIME, Department of Physical Therapy, The placebo group received a placebo (simulated laser therapy).
Federal University of Juiz de Fora, R. Main Outcome Measures: The primary outcomes were collected at baseline, 4 weeks
Leonardo Cristino, 3400—São Geraldo,
Gov. Valadares—MG, Minas Gerais CEP and 8 weeks. Pain was assessed through visual analogue scale (VAS) and pressure
35012000, Brazil. pain thresholds (PPT). Bite force was collected by a load cell synchronised with sur‐
Email: [email protected]
face electromyography of masticatory muscles, bilaterally.
Funding information Results: Pain scores decreased for both groups, but the intervention group showed
Fundação de Amparo à Pesquisa do
Estado de Minas Gerais; Coordenação lower values at 8 weeks. No differences were noted between groups for PPT, but the
de Aperfeiçoamento de Pessoal de Nível results increased for both overtime. Time until fatigue and muscle efficiency were
Superior
higher in the intervention group vs placebo group in both within‐ and between‐sub‐
ject analysis. Force increased from 4 to 8 weeks in the PG, without differences be‐
tween groups. Temporal muscle excitation was higher on 8 weeks compared with
baseline for the intervention group, without differences between groups.
Conclusion: Eight‐week exercise protocol of muscle endurance alleviates the pain
and improves the resistance to fatigue and muscle efficiency in TMD subjects.

KEYWORDS
bite force, electromyography, exercise therapy, facial pain, resistance training,
temporomandibular joint disorders

J Oral Rehabil. 2019;46:885–894. wileyonlinelibrary.com/journal/joor

© 2019 John Wiley & Sons Ltd | 885
886 | BARBOSA et al.

1 | I NTRO D U C TI O N until fatigue and the perceived pain (algometry and visual analogue
scale) on masticatory muscles.
The prevalence of temporomandibular disorder (TMD) is 27%‐38%
of the examined adult population.1 Women showed higher preva‐
2 | M E TH O DS
lence of painful conditions than men, including both oro‐facial pain
and other TMD symptoms. 2-4 Subjects with TMD present more
2.1 | Material
easily fatigued and less efficient masticatory muscles, with lower
oxygen extraction capacity during mastication than healthy control The subjects (Figure 1) were recruited by public invitation through
subjects.5 Such condition leads to reduced supply of oxygen to the folders and personal contacts; the authors assume that this could
6
muscle that interferes with the contractile function. Additionally, represent a selection bias. However, there is no public or private
the severity of TMD and oro‐facial myofunctional disorders is re‐ service in the city to perform any TMD assessment, which im‐
lated to the oxygen extraction capacity percentage.6 pairs the voluntary seek for treatment. A sample of 98 women
Conservative interventions for subjects with TMD including ex‐ voluntarily presented themselves for the study. The Research
ercises, joint mobilisation, splints and combinations of some of these Diagnostic Criteria for Temporomandibular Disorders (RDC/
techniques are often applied as primary choice for care.7,8 Low to TMD‐Axis I) was used for the diagnosis of TMD. The RDC/TMD
moderate level of evidence shows positive effects for improving is the internationally accepted gold standard, and its last version
symptoms of muscular TMD when using posture correction exer‐ is the DC/TMD. However, the lack of a validated Portuguese ver‐
9
cises for patients with myofascial pain. There was also a trend to fa‐ sion of DC/TMD led the authors to use the validated Portuguese
vour exercise therapy (general jaw exercises alone or combined with version of RDC/TMD for diagnostic purposes. Only those with
neck exercises in myogenous TMD) for pain free maximum opening chronic TMD (more than 6 months of complaints) were included.
and pain intensity when compared with a control group, with a mod‐ The jaw opening capacity was preserved at normal levels for
9
erate pooled effect size. However, the exercise programmes were all patients (above 4 cm), measured with a caliper, during RDC/
not often assessed alone, but in association to other therapies as TMD assessment. All subjects were diagnosed as myalgia, also
part of a conservative treatment protocol.7,9 A review with meta‐ according to RDC/TMD. The inclusion criteria for both groups
analysis analysed the effectiveness of manual therapy and thera‐ were to have a minimum of 28 permanent teeth and age between
peutic exercise for TMD.9 The authors suggested that more trials 18 and 45 years old. All subjects reported no periodontal issues.
isolating the type of exercise are necessary, mentioning that further Exclusion criteria for both groups were history of trauma on the
research is required to assess the usefulness of aerobic exercise as face and on the temporomandibular joint (TMJ), systemic dis‐
well as focused muscular training. Exercise therapy also shows in‐ eases such as arthritis, pain attributable to confirmed migraine,
consistent results due to the lack of appropriate dosage parameters head, or neck pain condition, chronic use (more than 6 months)
and the consequent failure to identify the effectiveness of exercise of any analgesic, anti‐inflammatory or psychiatric drugs, acute
prescription.7,9 infection or other significant disease of the teeth, ears, eyes,
Lengthy duration of submaximal contractile activity during en‐ nose, or throat, and to present neurological or cognitive deficit.
durance exercises requires a continuous supply of energy, provided After the first screening, only 46 subjects with TMD associated
by the local storage and through increased blood flow to skeletal with muscular oro‐facial pain were included and randomly di‐
muscles.10 The muscle ability to produce ATP for movement occurs vided by an independent rater into two groups: placebo group
by adapting the mRNA levels and protein changes to increase mito‐ (n = 23), and intervention group (n = 23). The randomisation was
chondrial concentrations as the endurance exercise progresses.10,11 carried out by an independent rater considering the 1:1 alloca‐
The increased mitochondrial concentrations are not the only adap‐ tion ratio. Before the study begins, a random allocation sequence
tations to endurance training. Energetic demands of muscle con‐ was automatically generated using the Research Randomizer
traction due to such exercise are also controllers of glucose ratio website (www.rando​m izer.org), by using 1 set of numbers, with
and whole‐body fat levels.10,12,13 TMD subjects have significantly a total of 46 numbers per set, and the established number range
lower endurance time during functional biting activities (such as as 1‐2, representing the placebo and the intervention group, re‐
bilateral of unilateral chewing) compared with controls.14 However, spectively. The random sequence was delivered by the Research
the response of the jaw muscles to local endurance exercise remains Randomizer, and the independent rater kept the sequence. The
poorly known despite the benefits of endurance exercises for other sequence order was continuously given to the examiner who per‐
skeletal muscles. formed the assessments when a new participant was allocated
The present study aimed to assess the effects of an 8‐week pro‐ for treatment. The rater who performed the randomisation was
tocol of local endurance exercises. The hypothesis is that the 8‐week blinded to the statistical analysis. The RDC/TMD assessments,
protocol of local endurance exercises would change the biomechani‐ the physical examination, comprising muscle and TMJ palpation,
cal parameters (time until fatigue, eletromyographic excitation, force measurement of active mandibular movements and joint noise
response and muscle efficiency considering the force‐time param‐ analysis were performed by well‐trained professionals. The ran‐
eters) of the masseter and temporal muscles during maximal biting dom allocation list was not accessible to the recruiting staff or
BARBOSA et al. | 887

FIGURE 1 Flow diagram

to the physiotherapists who implemented the treatment at any The subjects were assessed before (baseline), at 4 weeks and at
time. The group allocation and the allocation concealment were 8 weeks after the protocol begins. The a priori sample size calculation
preserved. Personal reasons (travel during more than a week and was based on a previous study,17 considering the effect size of 0.60, the
lack of time for treatment) lead six subjects from each group to alpha level of 5% and a 95% power, returning a total of 32 individuals.
discontinue the protocol. Thirty‐four were analysed at 8‐week as‐
sessments. The Mandibular Function Impairment Questionnaire
2.2 | Instruments
was used to classify the subjects in relation to the severity of the
functional limitation related to TMD (Table 1).15,16 The baseline Surface electromyography was used to evaluate the muscle excita‐
participant's characteristics were reported in Table 1. The eth‐ tion of the temporal and masseter muscles during a fatigue biting
ics committee for human investigation of the Federal University task for both Placebo and Intervention groups. The continuous biting
of Juiz de Fora approved the procedures employed in the study task leads to extreme fatigue, and it was performed only once. An
(protocol number 68457617.6.0000.5147). The objectives of the acquisition module with eight analog channels (Miotec™, Biomedical
study were explained to the subjects, and they were notified of Equipments) continuously recorded the biological signals. The conver‐
the benefits and potential risks involved before signing an in‐ sion from analog to digital signals was performed by an A/D board
formed consent form prior to the participation. This study was with 14‐bit resolution input range, the sampling frequency of 2 kHz,
registered in the Brazilian clinical trials registry (ensaiosclinicos. common rejection module greater than 100 dB, signal‐noise ratio
gov.br): protocol number RBR‐6kyh2g. less than 03 μV Root Mean Square and impedance of 109 Ω. The
888 | BARBOSA et al.

TA B L E 1 Subjects' characteristics
2.4 | Fatiguing biting test
Characteristic Control Intervention P
The fatigue of the masseter and temporalis muscles was assessed
n 17 17 –
during a single maximal biting effort test. The previously described
Age (y) 26 ± 8 30 ± 7 0.13a MVIC procedures were adopted. However, instead of an immediate
Weight (kg) 58 ± 9 64 ± 14 0.14a short maximal effort, each participant was asked to perform a MVIC
Height (cm) 162 ± 5 164 ± 7 0.44a for the maximal supported time.
BMI (kg/m2) 22 ± 4 24 ± 4 0.17a
Severity indexc
2.5 | Pain assessments
Low 11 (64.7%) 9 (52.9%) 0.39b
Moderate 6 (35.3%) 8 (47.1%) A visual analog scale (VAS) of 0‐100 mm, which was designed as 0

a
being no pain and 100 being the worst pain ever experienced, was
Independent t test.
b
Chi‐square binomial test.
given to subjects in this study to mark a perpendicular line between the
c
The severity index was obtained using the Mandibular Function 2 extremes to represent their pain intensity at the time of evaluation
Impairment Questionnaire. with provoked pain through temporal palpation.18 Psychometric prop‐
erties of VAS were previously tested with excellent results to evalu‐
collected data were windowed at 125 ms using the Miotec™ Suite ate pain.19-21 All VAS scores were collected by the same rater. Visual
Software. The sEMG signals were recorded in root mean square in analog scale scores ⩽3.4 represent mild pain, 3.5‐7.4 moderate pain
μV with surface Meditrace™ (Ludlow Technical Products) Ag/AgCl and ⩾7.5 severe pain.22 The VAS results were used for comparisons.
electrodes with a diameter of 1 cm and centre‐to‐centre distance of The pressure pain threshold (PPT) was measured with the de‐
1 cm, applied in a transverse orientation parallel to the underlying fi‐ scribed load cell adapted as an algometer (Figure S2). The equip‐
bres on a muscle site. A reference electrode was placed on the left ment's intra‐rater reliability was previously tested with 1 week apart
lateral humeral epicondyle. sEMG signals were amplified and filtered repeated measures. The results were adequate, with an ICC of 0.83.
(Butterworth fourth‐order, 20‐450 Hz bandpass filter, 60 Hz notch The load cell was recalibrated for each participant following the man‐
filter). All pieces of information were recorded and processed using ufacturer recommendations. All PPT measurements were performed
the software Miotec Suite™ (Miotec Biomedical Equipments). Prior to by the same investigator in the following order: left temporal, left
the sEMG electrode placement, the skin was cleaned with 70% alco‐ TMJ, left masseter, right temporal, right TMJ and right masseter. As
hol to eliminate residual fat, followed by an exfoliation using a specific the signal was collected continuously, the previous order was always
sandpaper for skin and a second cleaning with alcohol. The electrodes the same to identify each PPT. The volunteer remained seated with
were positioned on the anterior temporal muscles and the superficial the trunk erect, feet on the floor and hands resting on the thighs. The
masseter on both left and right sides parallel to the muscle fibres.3 device has a 1‐cm2 rubber application head, and measurements were
calculated as kg/cm2. Analog to digital conversion was performed
by the previously described acquisition module. The PPT was mea‐
2.3 | Maximal voluntary isometric contraction
sured bilaterally over the TMJ (precisely in front of the ear canal), the
Three 5‐second maximum isometric contraction (MVIC) were per‐ belly of the masseter muscles and the belly of the anterior temporal
formed by each participant while biting on an adapted load cell muscles.23 These sites received progressive 1 kg/s pressure until the
(Miotec™, Biomedical Equipments; maximum tension‐compres‐ participant experienced pain. The participant lifted a hand when the
sion = 200 kgf, precision of 0.1 kgf, maximum error of measure‐ PPT was achieved. The software continuously recorded all measures
ment = 0.33%). Each MVIC was followed by 5 minutes of rest. for offline comparisons. The PPT was measured three times at each
Subjects were asked to sit comfortably (the volunteer remained site with 3‐second interval. All assessments were collected in a sep‐
seated with the trunk erect, feet on the floor and hands resting on arate session from the exercise session.
the thighs) while the load cell arms were positioned on the incisors
(Figure S1). A disposable material was used to cover the arms for
2.6 | Intervention
each subject. The forward head posture was controlled during all
procedures by positioning the load cell closer to the participant, so The protocol consisted of resistance exercises twice a week for
the subjects could bite in their natural head posture. Standardised 8 weeks (16 sessions). Table 2 shows the progressive protocol de‐
verbal commands (“start,” “keep biting,” “stop”) were used by the veloped to respect the principle of low external load, allowing more
same rater for all tests' recordings. A 5‐second familiarisation was series and repetitions. The external load ranged from 20% to 50%
followed by 3 minutes of rest before the MVIC. The intra‐rater relia‐ of the MVIC, and it was controlled using the Biotrainer™ visual bio‐
bility of the adapted load cell was previously assessed twice (1 week feedback software (Miotec™, Biomedical Equipments – Figure 2).
apart measures), returning a very good ICC of 0.84. The load cell was The protocol was previously set by a trained rater, using the same in‐
coupled and synchronised with the electromyography. The mean structions, but no verbal encouragement was given. The participant
among the MVIC trials was used for normalisation purposes. was instructed to perform short bites on the adapted load cell, and
BARBOSA et al. | 889

TA B L E 2 Intervention protocol

Rest before Time of Interval between Rest between Total time of

Week contraction contraction Load % of MVIC Repetitions repetitions Series series execution/session

1 7s 5s 20% 10 2s 3 2 min 7 min 41 s

2 6s 5s 20% 12 2s 4 2 min 11 min 5 s
3 7s 4s 25% 15 2s 4 1.5 min 10 min 39 s
4 8s 4s 30% 18 2s 4 1.5 min 11 min 51 s
5 9s 3s 35% 20 1s 5 1 min 10 min 50 s
6 10 s 3s 40% 22 1s 5 1 min 11 min 9 s
7 11 s 2s 45% 25 1s 5 0.5 min 8 min 25 s
8 12 s 2s 50% 25 1s 5 0.5 min 8 min 25 s

an additional familiarisation session was allowed at the 1st week.

The external load, the repetitions, the rest before contraction and
the series were progressively increased, while the time of contrac‐
tion, the interval between repetitions and the rest between series
were progressively decreased.
The other group received placebo via simulated low intensity
laser therapy (off mode) for the same time of session as the inter‐
vention group protocol (the off mode enabled the equipment to
emit beep sounds, but without laser application). The simulated laser
therapy followed all phases of eye protection (special glasses for the
patient and for the therapist) and equipment's positioning. The sites
of positioning were as follows: the TMJ, the anterior temporal mus‐
cle and the masseter muscle, bilaterally.
F I G U R E 2 Electromyographic signal obtained from the
Biotrainer™ biofeedback interface by using the load cell to
2.6.1 | Raters control the amplitude of the contraction. A, Rest period; B,
contraction period; C, inter‐contraction rest period; D, the grey
An independent rater (rater 1) performed the randomisation pro‐
rectangle corresponds to the amplitude of contraction preset by
cedure, as explained. Three raters (raters 2, 3 and 4) were ex‐ the therapist. The subject was instructed to sustain the level of
haustively trained for 6 months before the study to perform the contraction by reaching the superior level of the rectangle and then
assessments (muscle palpation, pain assessments and how to rest during the interval between rectangles
perform the RDC/TMD) and to apply both treatments. The raters
were not allowed to perform both (assessments and treatments). muscle excitation, as the exercise may affect the motor unit recruit‐
The long‐term training was necessary to ensure the procedures' re‐ ment before structural tissues changes; and (c) pain perception (VAS
liability and included how to behave near the subjects to minimise scores and PPT).
bias of subjects' perceptions. The calibration was first performed
by one with prior experience with the RDC/TMD (gold standard
2.8 | Data extraction
examiner—rater 2). Two remaining examiners (trained examiners—
raters 3 and 4) were trained and calibrated by the first, in addition All data were offline extracted using the Miotec Suite™ Software
to watching the RDC/TMD exam training video (available at http:// (Miotec™, Biomedical Equipments). As the load cell was synchro‐
www.rdc-tmdin​terna​tional.org). Data extraction and the statistical nised with the electromyography channels, the trained rater set the
analysis were performed by the independent rater (rater 1). interval using the force onset. After three 1‐second windows of rest
were collected, the onset was defined by three times the standard
deviation from the averaged rest intervals plus the mean itself. The
2.7 | Primary and secondary outcomes
interval started when the signal exceeded the onset threshold value.
The primary outcome was the time until fatigue during the fatigu‐ Conversely, the end of the interval was set using the same thresh‐
ing biting test, as the main goal of the protocol was to improve the old. Interval means were used for statistical analysis (force, muscle
muscle endurance. Secondary outcomes were as follows: (a) effects excitation). The total time until fatigue (from the onset until the end
on muscle efficiency, representing a ratio between the time until fa‐ of the interval) was also collected from the software recordings.
tigue and the generated force during the biting task; (b) effects on The VAS scores and the mean from the three peaks of algometry
890 | BARBOSA et al.

were considered for the statistical analysis. The muscle efficiency without significant differences are summarised at Table S1. The
considered the mean total of force times the amount of time spent subjects were treated in the school clinic of Governador Valadares
to perform the task. The result was divided by 100 to be expressed (Minas Gerais, Brazil) between January 2018 and December 2018.
as a percentage. All treatments were provided onsite. No practice at home was asked.
Attendance at sessions was, therefore, taken as compliance with the
treatment protocol. No co‐interventions were performed in either
2.9 | Statistical analysis
group, and no adverse effects were reported by any participant dur‐
Data were presented as means and standard deviation. The multi‐ ing the 8‐week protocol. Thirty‐four subjects finished the protocol.
variate analysis of variance general linear model with repeated meas‐
ures was used to rate differences within and between groups and
3.1 | Composite variable comparisons
to extend the analysis by taking into account multiple continuous
dependent variables, bundling them together into a weighted linear Significant differences were noted between groups (F = 2.4; P = 0.04),
combination or composite variable (Hotelling's trace). Sidak's post and within‐subjects analysis showed significant differences for time
hoc test was used to perform pairwise comparisons. The significance effect (F = 3.7; P = 0.0001) on the composite variable. However, no
was set at P < 0.05. All analysis was done using the SPSS Inc (PASW differences were noted considering the interaction time‐by‐group
Statistics for Windows, version 18.0. Chicago: SPSS Inc) and G‐Power when all variables were bundled together into a composite (F = 1.5;
software (version 3.1.5, Franz Faul, Universität Kiel, Germany). P = 0.08).

3.2 | Between‐group comparisons

3 | R E S U LT S
The between‐group pairwise comparisons showed differences for
Table 3 summarises the descriptive and inferential data from vari‐ the VAS (F = 4.05; P = 0.04) at 8 weeks. The intervention group
ables with significant differences. The descriptive data of variables showed lower value compared with the placebo group. Differences

TA B L E 3 Variables with significant differences

Post hoc be‐

tween‐group
Baseline 4 wk 8 wk Post hoc within‐group P (95% CI);
Outcomes Groups Mean (SD) Mean (SD) Mean (SD) P (95% CI) Moment

VAS (cm)* Intervention 3.7 (2.2) 2.6 (2) 1.1 (1.3) 0.003 (0.8; 4) 0.01 (0.4; 3.3);
8 wk < Baseline 8 wk
0.001 (0.7; 2)
8 wk < 4 wk
Placebo 4.9 (3.4) 3.6 (2.5) 3.0 (2.6) 0.03 (0.3; 4)
8 wk < Baseline
Efficiency** Intervention 37 (23) 56 (31) 68 (36) 0.02 (2; 34) 0.01 (6; 43); 4 wk
4 wk > Baseline 0.001 (14; 53);
0.03 (1; 24) 8 wk
8 wk > 4 wk
0.0001 (16; 47)
8 wk > Baseline
Placebo 26 (13) 32 (21) 35 (14) NS
Time until fatigue Intervention 42 (21) 65 (40) 74 (36) 0.02 (2; 43) 0.003 (12; 54);
(s)*** 4 wk > Baseline 8 wk
0.001 (11; 52)
8 wk > Baseline
Placebo 35 (18) 41 (23) 40 (22) NS
EMG TEMP‐R Intervention 75 (21) 83 (37) 113 (51) 0.007 (9; 67) –
(%)**** 8 wk > Baseline
Placebo 86 (39) 81 (35) 76 (49) –

Note: Descriptive data in mean and SD. Within‐ and between‐subject post hoc comparisons.
Significant within‐group differences:
*F = 7.2; P = 0.001 at 8 wk;
**F = 7.4; P = 0.01 at 4 wk and F = 12.8; P = 0.001 at 8 wk;
***F = 4.5; P = 0.04 at 4 wk and F = 10.6; P = 0.003 at 8 wk;
****F = 4.7; P = 0.04 at 8 wk.
BARBOSA et al. | 891

were observed between placebo and intervention group for ef‐ the expected pain levels and emotional feelings, such as reduced
ficiency at 4‐week and 8‐week assessments (F = 11.15; P = 0.002), anxiety and the previous experience of relief.31,32 The current study
with progressive higher values of efficiency on the intervention used a simulated low intensity laser therapy with the equipment
group. The time until fatigue also showed significant differences at emitting beep sounds, but without the actual laser application. The
8 weeks (F = 8.25; P = 0.007). The intervention group performed the equipment is widely used by physical therapists to treat pain and
biting task for much more time than the placebo group. inflammatory musculoskeletal conditions. As the simulated laser
therapy followed all phases of eye protection and equipment's po‐
sitioning on the sites of pain, the subject's experience and possible
3.3 | Within‐group comparisons
expectations of relief were possible affected, inducing a lower pain
Considering the time factor, within‐group comparisons for VAS perception. However, the effect was lower than the exercise at the
showed a progressive decrease in perceived pain for both groups. end of the protocol and limited to a certain level without differences
Significant differences in the placebo group occurred between the between 4‐ and 8‐week assessments.
baseline and the 8‐week evaluations, with lower values at the final There are many different exercises protocols and the outcome
assessment. Variations were also observed on the intervention may vary with the prescription, especially for TMD.7,9 The external
group for VAS between the baseline and 8 weeks, and between the load, the number of series and repetitions, and the rest between se‐
4‐ and 8‐week assessments, with progressive lower values of VAS. ries and the training frequency are the main factors controlled by the
No discrepancy was noted overtime for placebo group on efficiency physical therapist.33 Also, the many possible combinations of these
or time until fatigue. Higher time until fatigue and muscle efficiency factors and individual metabolic response could influence some of
were progressively observed at 4‐ and 8‐week ratings for the inter‐ the results associated with a specific exercise protocol. The authors
vention group. No within‐group differences were observed for the acknowledge that there are several difficulties to control exercise
force response, the algometry and the muscle excitation. variables in clinical and laboratory settings. Equipment with bio‐
feedback to control external load exercises is usually expensive, not
adaptable for biting, and the software are not always friendly user.
4 | D I S CU S S I O N However, without controlling the exercise parameters, all inferences
would be biased. The choice of adapting the load cell allowed both
This study assessed the effects of an 8‐week protocol of local en‐ assessment and training.
durance exercises on masticatory muscles in women with TMD. Individualised, supervised exercise based on patient presenta‐
The prevalence of TMD symptoms ranges from 15% to 35%. 24,25 tion and preferences is essential for controlling chronic pain.34,35
However, as the sample was stimulated to seek assessment and Though, despite such individual variety, the overall response among
treatment, this study screened 98 subjects finding 61% of those with subjects to an exercise protocol tends to remain similar to a group
positive diagnosis of TMD. Additionally, TMD symptoms tend to be when more physical and even phycological similarities occur.33,36,37
underdiagnosed due to similar complaints of other disorders, such The present study particularly focused on women with oro‐facial
as headaches, fibromyalgia, and painful cervical and shoulder. 26,27 pain and TMD during the most prevalent phase of life for these
The present study also had 26% dropout rate, while another ran‐ conditions. Relevant variables changed overtime due to the exer‐
domised controlled study with combined therapy (laser + exercises) cise protocol, partially confirming the hypothesis. At the end of the
had 20%. 28 A study with the same design using stabilisation splints protocol, the VAS scores were lower for the intervention compared
in TMD subjects had 33% dropout rate. 29 TMD subjects are used with placebo group. Other studies already have shown dissimilarity
to receive a multimodal conservative approach. As the exercise was in pain due to exercises in TMD subjects. A prospective study in‐
applied alone, the present dropout rate was probably due to a sin‐ cluded coordination, endurance and strengthening exercises for the
gle type of therapy. The choice to use the exercise without other jaw‐neck‐shoulder region in an individualised 10‐ to 24‐week proto‐
types of treatment was to ensure the internal validity. Perceived col. All subjects reported a reduction in jaw pain after the exercise
pain scores decreased overtime for both groups, but the interven‐ programme, classified through a numerical rating scale. 38 Another
tion group showed lower values at 8 weeks compared with the pla‐ study consisted of 10 sessions of muscle‐conditioning techniques,
cebo group. The time until fatigue and the muscle efficiency were manual therapy and stretching over 5 weeks in 12 women with
higher on the intervention group in both within‐ and between‐group mixed TMD (combining myofascial pain either with joint impair‐
comparisons. Temporal muscle excitation was higher on 8‐week ment or disc displacement).16 The results showed significant man‐
evaluation compared with baseline on the intervention group, but dibular function improvement and decreased self‐reported pain
no between‐group divergences were observed. score. Although systematic reviews and meta‐analysis observed
Lower values of VAS on the placebo group were expected, as this two major issues on the majority of studies involving exercises,
effect is derived from the participants' perception and experience they did not report interventions sufficiently to be reproducible,
of receiving a pain‐reducing treatment as well as the integration of and co‐interventions were also not controlled.7,9 The current study
this sensory information with memories of previous experiences and isolated the exercise intervention and controlled all parameters in‐
current expectations.30,31 Placebo effects are also associated with volved in the prescription, with relevant changes in pain. A recent
892 | BARBOSA et al.

review suggested peripheral and central nervous system sensitisa‐ from the results of this study. The sample were not subjects seek‐
tion due to mechanical pressure pain sensitivity in the trigeminal ing treatment, and this could represent a selection bias. The pain
region and remote regions in subjects with TMD. 39 The same study was provoked by palpation without report of familiar or spontane‐
suggested spinal and central hyperexcitability in TMD subjects. The ous pain. Psychosocial assessments may influence the group split.
central sensitisation process may amplify the pain information in Despite sample size calculation, the number of subjects who met the
the brain, resulting in a reduction in the normal central inhibitory eligibility criteria was relatively restricted.
mechanisms that help to balance activation of pain centres.40 The
TMJ pain is an usual complaint in subjects with TMD, and it can be
a referred pain from the myalgia due to central or even peripheral 5 | CO N C LU S I O N S
23,41,42
sensitisation. The PPT increased overtime for both groups,
but no differences were observed within or between groups any‐ Physical rehabilitation with exercise protocol focusing in local mus‐
time. Objectively, the placebo was as good as the exercise protocol. cle endurance training alleviates the perceived oro‐facial pain and
A possible explanation is that the present exercise approach and improves the fatigue and the muscle efficiency in TMD subjects.
assessments to TMD pain did not take into account for the psycho‐ The control of exercise parameters with a biofeedback system was
social sphere, mainly due to excessive number of variables and time important to establish an objective clinical progression. The local en‐
consumed to evaluate during the sessions. The psychosocial factors durance training was clinically relevant to treat TMD, improving the
have relevant role in both classifying and treating TMD subjects.43 fatigue threshold. Further research might focus on affordable equip‐
Subjects with emotional profile with low disability, high intensity ment to control external loads during exercises for TMD.
pain‐related impairment, and high to moderate levels of somatisa‐
tion and depression would be important split factors or a co‐vari‐
AC K N OW L E D G M E N T S
ates to include in future assessments.43
The higher time until fatigue in the intervention group compared Special thanks to UFJF‐GV Department of Physical Therapy and to
with placebo was the main factor to change the muscle efficiency the Diretoria de Relações Internacionais—UFJF. This study was fi‐
as the force remained the same between groups across time. The nanced in part by the Coordenação de Aperfeiçoamento de Pessoal
unchanged biting force was expected due to the characteristics de Nível Superior—Brasil (CAPES)—Finance Code 001, and by the
of the exercise protocol. High external load (>60% of MVIC) and Fundação de Amparo à Pesquisa de Minas Gerais (FAPEMIG).
low repetitions (6‐12) are essential to strength changes.13 But for
local muscular endurance training, it is recommended that light to
ORCID
moderate loads (40%‐60% of MVIC) be performed for high repeti‐
tions (>15) using short rest periods (<90 seconds).13 The previous Alexandre Carvalho Barbosa https://orcid.
studies observed a relationship among fatigue, masticatory mus‐ org/0000-0001-7862-1737
cles efficiency and TMD. 6,44 Subjects with TMD showed reduced
endurance to jaw motor tasks with lower oxygen extraction capac‐
ity compared with healthy subjects. 5,6,44 Individuals with impaired REFERENCES

functional capacity with regard to endurance may have increased 1. Anastassaki Köhler A, Hugoson A, Magnusson T. Prevalence of
risk for developing pain and dysfunction.14,38 Nevertheless, gen‐ symptoms indicative of temporomandibular disorders in adults:
eral aerobic exercises have shown to improve muscle strength, cross‐sectional epidemiological investigations covering two de‐
cades. Acta Odontol Scand. 2012;70(3):213‐223.
flexibility, and functional capacity and could induce analgesia.45
2. Ramalho D, Macedo L, Goffredo Filho G, Goes C, Tesch R.
The current results showed that the level of perceived pain de‐ Correlation between the levels of non‐specific physical symp‐
creased as the time until fatigue increased, and consequently, toms and pressure pain thresholds measured by algometry
the muscle efficiency was improved without changes in force re‐ in patients with temporomandibular disorders. J Oral Rehabil.
2015;42(2):120‐126.
sponse. These novel findings highlight the local muscle endurance
3. Woźniak K, Lipski M, Lichota D, Szyszka‐Sommerfeld L. Muscle fa‐
exercises as an alternative to treat TMD subjects with oro‐facial tigue in the temporal and masseter muscles in patients with tem‐
pain. poromandibular dysfunction. Biomed Res Int. 2015;2015:1‐6.
4. Bueno CH, Pereira DD, Pattussi MP, Grossi PK, Grossi ML. Gender
differences in temporomandibular disorders in adult populational
4.1 | Study limitations studies: a systematic review and meta‐analysis. J Oral Rehabil.
2018;45(9):720‐729.
Further studies comparing combinations among other successful 5. Xu L, Fan S, Cai B, Fang Z, Jiang X. Influence of sustained sub‐
therapies are needed to provide the best care for TMD subjects. maximal clenching fatigue test on electromyographic activ‐
The present study also focused on women at limited range of age. ity and maximum voluntary bite forces in healthy subjects and
patients with temporomandibular disorders. J Oral Rehabil.
Those presenting other levels of functional limitations and other
2017;44(5):340‐346.
age groups may show distinct patterns, as their male counterpart. 6. Ferreira C, Bellistri G, Montagna S, de Felício CM, Sforza C. Patients
The placebo procedure may also influence pain results, as inferred with myogenic temporomandibular disorders have reduced
BARBOSA et al. | 893

oxygen extraction in the masseter muscle. Clin Oral Investig. I and II) for temporomandibular disorders (the Maringá study). Int J
2017;21(5):1509‐1518. Prosthodont. 2015;28:600‐609.
7. Dickerson SM, Weaver JM, Boyson AN, et al. The effectiveness of 25. Jussila P, Kiviahde H, Näpänkangas R, et al. Prevalence of temporo‐
exercise therapy for temporomandibular dysfunction: a systematic mandibular disorders in the Northern Finland Birth Cohort 1966. J
review and meta‐analysis. Clin Rehabil. 2017;31(8):1039‐1048. Oral Facial Pain Headache. 2017;31(2):159‐164.
8. Shaffer SM, Brismée J‐M, Sizer PS, Courtney CA. 26. Ferreira C, da Silva M, de Felício CM, Ferreira C, da Silva M, de
Temporomandibular disorders. Part 2: conservative management. Felício CM. Sinais e sintomas de desordem temporomandibular em
J Man Manip Ther. 2014;22(1):13‐23. mulheres e homens. CoDAS. 2016;28(1):17‐21.
9. Armijo‐Olivo S, Pitance L, Singh V, Neto F, Thie N, Michelotti A. 27. Lupoli TA, Lockey RF. Temporomandibular dysfunction: an often
Effectiveness of manual therapy and therapeutic exercise for tem‐ overlooked cause of chronic headaches. Ann Allergy, Asthma
poromandibular disorders: systematic review and meta‐analysis. Immunol. 2007;99(4):314‐318.
Phys Ther. 2016;96(1):9‐25. 28. da Silva MM, Albertini R, Leal‐Junior EC, et al. Effects of exercise
10. Booth FW, Ruegsegger GN, Toedebusch RG, Yan Z. Endurance ex‐ training and photobiomodulation therapy (EXTRAPHOTO) on pain
ercise and the regulation of skeletal muscle metabolism. Prog Mol in women with fibromyalgia and temporomandibular disorder: study
Biol Transl Sci. 2015;135:129‐151. protocol for a randomized controlled trial. Trials. 2015;16(1):252.
11. Perry C, Lally J, Holloway GP, Heigenhauser G, Bonen A, Spriet LL. 29. Pho Duc JM, Hüning SV, Grossi ML. Parallel randomized controlled
Repeated transient mRNA bursts precede increases in transcrip‐ clinical trial in patients with temporomandibular disorders treated
tional and mitochondrial proteins during training in human skeletal with a CAD/CAM versus a conventional stabilization splint. Int J
muscle. J Physiol. 2010;588(Pt 23):4795‐4810. Prosthodont. 2016;29(4):340‐350.
12. Ramírez‐Campillo R, Andrade DC, Campos‐Jara C, Henríquez‐ 30. Price DD, Finniss DG, Benedetti F. A comprehensive review of the
Olguín C, Alvarez‐Lepín C, Izquierdo M. Regional fat changes in‐ placebo effect: recent advances and current thought. Annu Rev
duced by localized muscle endurance resistance training. J Strength Psychol. 2008;59(1):565‐590.
Cond Res. 2013;27(8):2219‐2224. 31. Vase L, Skyt I, Hall KT. Placebo, nocebo, and neuropathic pain. Pain.
13. American College of Sports Medicine. Progression models 2016;157:S98‐S105.
in resistance training for healthy adults. Med Sci Sport Exerc. 32. Piedimonte A, Guerra G, Vighetti S, Carlino E. Measuring expecta‐
2009;41(3):687‐708. tion of pain: contingent negative variation in placebo and nocebo
14. Wänman A. Endurance to physical strain in patients with temporo‐ effects. Eur J Pain. 2017;21(5):874‐885.
mandibular disorders: a case‐control study. Acta Odontol Scand. 33. Garber CE, Blissmer B, Deschenes MR, et al. Quantity and quality
2012;70(6):455‐462. of exercise for developing and maintaining cardiorespiratory, mus‐
15. Campos J, Carrascosa AC, Maroco J. Validity and reliability of culoskeletal, and neuromotor fitness in apparently healthy adults.
the Portuguese version of Mandibular Function Impairment Med Sci Sport Exerc. 2011;43(7):1334‐1359.
Questionnaire. J Oral Rehabil. 2012;39(5):377‐383. 34. Booth J, Moseley GL, Schiltenwolf M, Cashin A, Davies M, Hübscher
16. Calixtre LB, Grüninger BLdS, Haik MN, Alburquerque‐sendín M. Exercise for chronic musculoskeletal pain: a biopsychosocial ap‐
F, Oliveira AB. Effects of cervical mobilization and exercise on proach. Musculoskeletal Care. 2017;15(4):413‐421.
pain, movement and function in subjects with temporoman‐ 35. Kroll HR. Exercise therapy for chronic pain. Phys Med Rehabil Clin N
dibular disorders: a single group pre‐post test. J Appl Oral Sci. Am. 2015;26(2):263‐281.
2016;24(3):188‐197. 36. Ishiyama H, Inukai S, Nishiyama A, et al. Effect of jaw‐opening exer‐
17. Machado B, Mazzetto MO, Da Silva M, de Felício CM. Effects of cise on prevention of temporomandibular disorders pain associated
oral motor exercises and laser therapy on chronic temporomandib‐ with oral appliance therapy in obstructive sleep apnea patients: a
ular disorders: a randomized study with follow‐up. Lasers Med Sci. randomized, double‐blind, placebo‐controlled trial. J Prosthodont
2016;31(5):945‐954. Res. 2017;61(3):259‐267.
18. Costa YM, Porporatti AL, Calderon P, Conti P, Bonjardim LR. Can 37. Bolgla LA, Earl‐Boehm J, Emery C, Hamstra‐Wright K, Ferber
palpation‐induced muscle pain pattern contribute to the differen‐ R. Pain, function, and strength outcomes for males and fe‐
tial diagnosis among temporomandibular disorders, primary head‐ males with patellofemoral pain who participate in either a hip/
aches phenotypes and possible bruxism? Med Oral Patol Oral Cir core‐ or knee‐based rehabilitation program. Int J Sports Phys Ther.
Bucal. 2016;21(1):e59‐e65. 2016;11(6):926‐935.
19. Herr KA, Spratt K, Mobily PR, Richardson G. Pain intensity assess‐ 38. Häggman‐Henrikson B, Wiesinger B, Wänman A. The effect of su‐
ment in older adults. Clin J Pain. 2004;20(4):207‐219. pervised exercise on localized TMD pain and TMD pain associated
20. Bird M‐L, Callisaya ML, Cannell J, et al. Accuracy, validity, and reli‐ with generalized pain. Acta Odontol Scand. 2018;76(1):6‐12.
ability of an electronic visual analog scale for pain on a touch screen 39. La Touche R, Paris‐Alemany A, Hidalgo‐Pérez A, López‐de‐Uralde‐
tablet in healthy older adults: a clinical trial. Interact J Med Res. Villanueva I, Angulo‐Diaz‐Parreño S, Muñoz‐García D. Evidence for
2016;5(1):e3. central sensitization in patients with temporomandibular disorders:
21. Carlsson AM. Assessment of chronic pain. I. Aspects of the reliabil‐ a systematic review and meta‐analysis of observational studies.
ity and validity of the visual analogue scale. Pain. 1983;16(1):87‐101. Pain Pract. 2018;18(3):388‐409.
22. Boonstra AM, Schiphorst Preuper HR, Balk GA, Stewart RE. Cut‐off 40. Harrison AL, Thorp JN, Ritzline PD. A proposed diagnostic clas‐
points for mild, moderate, and severe pain on the visual analogue sification of patients with temporomandibular disorders: implica‐
scale for pain in patients with chronic musculoskeletal pain. Pain. tions for physical therapists. J Orthop Sport Phys Ther. 2014;44(3):
2014;155(12):2545‐2550. 182‐197.
23. Herpich CM, Gomes CAFdP, Dibai‐Filho AV, Politti F, Souza C, 41. Cunha CO, Pinto‐Fiamengui L, Castro A, Lauris J, Conti P.
Biasotto‐Gonzalez DA. Correlation between severity of temporo‐ Determination of a pressure pain threshold cut‐off value for the
mandibular disorder, pain intensity, and pressure pain threshold. J diagnosis of temporomandibular joint arthralgia. J Oral Rehabil.
Manipulative Physiol Ther. 2018;41(1):47‐51. 2014;41(5):323‐329.
24. Progiante P, Pattussi M, Lawrence H, Goya S, Grossi P, Grossi M. 42. Etöz OA, Ataoğlu H. Evaluation of pain perception in patients
Prevalence of temporomandibular disorders in an adult brazilian with temporomandibular disorders. J Oral Maxillofac Surg.
community population using the research diagnostic criteria (axes 2007;65(12):2475‐2478.
894 | BARBOSA et al.

43. Canales G, Guarda‐Nardini L, Rizzatti‐Barbosa CM, Conti P,

Manfredini D. Distribution of depression, somatization and pain‐re‐ How to cite this article: Barbosa MA, Tahara AK, Ferreira IC,
lated impairment in patients with chronic temporomandibular dis‐ Intelangelo L, Barbosa AC. Effects of 8 weeks of masticatory
orders. J Appl Oral Sci. 2019;27:e20180210.
muscles focused endurance exercises on women with oro‐
44. Gavish A, Winocur E, Menashe S, Halachmi M, Eli I, Gazit E.
Experimental chewing in myofascial pain patients. J Orofac Pain. facial pain and temporomandibular disorders: A placebo
2002;16(1):22‐28. randomised controlled trial. J Oral Rehabil. 2019;46:885–894.
45. Sharma NK, Ryals JM, Gajewski BJ, Wright DE. Aerobic exercise https​://doi.org/10.1111/joor.12823​
alters analgesia and neurotrophin‐3 synthesis in an animal model of
chronic widespread pain. Phys Ther. 2010;90(5):714‐725.

S U P P O R T I N G I N FO R M AT I O N

Additional supporting information may be found online in the

Supporting Information section at the end of the article.