Journal of Oral Rehabilitation 2004 31; 287292

Occlusal adjustment for treating and preventing

temporomandibular joint disorders
H. KOH* & P. G. ROBINSON

*Department of Dental Public Health and Community Dental Education, GKT School of

Dentistry, London and Dental Public Health, School of Clinical Dentistry, Claremont Crescent, Sheffield, UK

SUMMARY To assess the effectiveness of occlusal

adjustment (OA) for treating temporomandibular
disorders (TMD) in adults and preventing TMD. The
Cochrane Controlled Trials Register, MEDLINE
and EMBASE were comprehensively searched using
the Cochrane methods. Reports and review articles
were retrieved. Unpublished reports or abstracts
were considered from the SIGLE database. All
randomized or quasi-randomized controlled trials
comparing OA with placebo, reassurance or no treatment in adults with TMD. The outcomes were global
measures of symptoms, pain, headache and limitation of movement. Data collection and analysis

Introduction
Temporomandibular disorder [TMD also known as
craniomandibular joint dysfunction, temporomandibular joint (TMJ) dysfunction] is a group of disorders of
the TMJ. The common signs and symptoms of TMD
include pain, joint sounds (clicking, grating), limited or
asymmetrical jaw movement and spasm of the chewing
muscles. These symptoms can have a profound effect
on health and quality of life (1).
It is a frequent problem (2). Prevalence studies have
reported approximately 75% of the population having
at least one sign of joint dysfunction (abnormal jaw
movement, joint noises, tenderness on palpation, etc.)
and approximately 33% having at least one symptom
(facial pain, joint pain, etc.) (3, 4). One other constellation of symptoms historically associated with TMD is
globus, which involves symptoms of choking sensations
or sore throat (5).
Various theories have been put forward that relate
the aetiology of TMD to organic disease of the TMJ,
2004 Blackwell Publishing Ltd

followed the Cochrane Oral Health Groups statistical guidelines. Results showed no difference
between OA and control group in symptom-based
outcomes for treatment or incidence of symptoms
for prevention. There is no evidence that OA treats
or prevents TMD. OA cannot be recommended for
the management or prevention of TMD. Future
trials should use standardized diagnostic criteria
and outcome measures when evaluating TMD.
KEYWORDS: occlusal adjustment, treatment, prevention, temporomandibular disorders
Accepted for publication 4 July 2003

trauma and stress (6). TMD has been associated with

occlusal factors for many years (7).
Treatment options for TMD include simple reassurance, physiotherapy (such as ultrasound, megapulse,
acupuncture, short wave diathermy laser, heat exercises and biofeedback), splint therapy, drug therapy,
surgical intervention and combined treatment (811).
Cochrane reviews of some of these treatments are
underway.
Occlusal adjustment (OA) is the selective adjustment of the teeth so that the upper and lower teeth
occlude harmoniously in the inter-cuspal position.
Adjustments can also be made to remove nonworking side contacts and contacts between the
posterior teeth when the mandible is moved anteriorly. OA has been used in the management of TMD
for many years but the evidence of effectiveness is
inconclusive (12). It is not clear if malocclusion has a
causal role in TMD. OA has been evaluated in studies
to prevent TMD (13) but again these studies were
inconclusive.

287

288

H. KOH & P. G. ROBINSON

Systematic reviews carefully search for, appraise and
synthesize primary data relating to a carefully focused research question. A Cochrane Review identifies
and, where possible, aggregates data from randomized
controlled trials (RCTs) of effectiveness of health care
interventions. They are particularly useful if the data of
effectiveness arises from small and/or contradictory
studies. Quantitative systematic reviews have not been
reported for treatments of TMD. One qualitative
systematic review has evaluated OA in treating TMD
(14) but did not include a meta-analysis of the existing
data. Therefore, this study aimed to establish the
effectiveness of OA in reducing symptoms in patients
with TMD and in preventing TMD.

Method
To be included in the review studies were to be
RCTs including quasi-randomized trials that assessed
OA in the management or prevention of TMD. Quasirandomized studies are studies where the method
of allocation was known but was not considered strictly
random. Non-randomized trials were excluded.
In trials of treatment, participants were to be adults
aged 18 years old or above with clinically diagnosed
TMD. The inclusion criteria required reports to state
their diagnostic criteria for TMD and for participants to
exhibit two or more of the signs and/or symptoms
(Table 1). There were no age restrictions for prevention
trials although prevention studies focused on children.
TMD was required to be clinically absent at baseline in
studies on prevention.
In each trial the treatment group was to have
received OA while the control group received no
treatment, placebo or reassurance. Studies where
splints had been used before treatment were excluded.
The interval required for outcome measurement was at
least 3 weeks after the intervention.
The primary outcomes were global symptoms, pain
and headache:
1 Relief from symptoms was assessed using global
measures of symptoms.
2 Data on pain were recorded according to frequency,
severity or duration. Where possible data for the
frequency, severity and duration of pain were aggregated using weighted mean differences but depended on
assessments of heterogeneity.
3 Similarly, data on headache were recorded according
to frequency, severity or duration. Where possible data

for the frequency, severity and duration of pain were

aggregated using weighted mean differences but depended on assessments of heterogeneity.
The secondary outcome considered was limitation of
movement. Other signs were ignored because they are
neither unique to the disease nor associated with the
progression or outcomes of TMD.
The search strategy for identification of studies was
based on the Cochrane Oral Health Group search
strategy. There was no language restriction for inclusion. Every effort was made to translate non-English
articles into English for inclusion.
The Cochrane Oral Health Group Specialized Register, Cochrane Controlled Trials Register (Issue 2, 2002),
MEDLINE (1966 to April 2002) and EMBASE (1980 to
April 2002) were searched electronically. The Cochrane
Controlled Trials Register was searched using the
strategy outlined in Table 2.
The following medical subject headings (MeSH)
terms were used in MEDLINE: temporomandibular
disorders; temporomandibular diseases; myofascial
pain; craniomandibular disorders; jaw joint problems.

Table 1. Inclusion criteria (23)

Participants should exhibit two or more of:
Symptoms
The occurrence of recurrent headache (two or more episodes
a month).
Pain in the jaws, face, throat, neck, shoulders or back.
Ear symptoms (includes tinnitus, stuffiness, diminished
hearing or pain).
Pain in the temporomandibular joint at rest and during
chewing.
Day- and night-time grinding or clenching.
Vertigo.
Stiffness in jaws.
Difficulties in swallowing.
Globus symptoms (associated with choking sensations or
soreness of the throat).
Joint sounds (including clicking and grating).
Spontaneous luxation or locking of the jaws.
Signs
Palpatory tenderness of the masticatory muscles.
Joint sounds during jaw movements.
Tenderness during jaw movements.
Deviation of the mandible on opening and closing.
Reduced mandibular range of motion.
Presence of occlusal interference in retruded, protruded and
medio- and latero-trusion positions of the mandible.
Wear facets.

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 287292

OCCLUSAL ADJUSTMENT FOR TREATING AND PREVENTING TMD

Table 2. Search strategy for the Cochrane Clinical Trials Register
#1 TEMPOROMANDIBULAR-JOINT*:ME
#2 (Explode) MYOFASCIAL-PAIN-DISORDERS*:ME
#3 (Explode) CRANIOMANDIBULAR-DISORDERS*:ME
#4 temporomandibular* OR temporo mandibular*
#5 craniomandibular* OR cranio mandibular*
#6 ((TMJ:TI or TMJ:AB) or TMJ:KY)
#7 ((CMD:TI or CMD:AB) or CMD:KY)
#8 ((TMD:TI or TMD:AB) or TMD:KY)
#9 (((((((#1 or #2) or #3) or #4) or #5) or #6) or #7) or #8)
#10 DENTAL-OCCLUSION-BALANCED*:ME
#11 OCCLUSAL-ADJUSTMENT*:ME
#12 ((occlus* NEAR balance*) OR (occlus* NEAR treat*) OR
(occlus* NEAR equilibrat*) OR (occlus* NEAR adjust*))
#13 ((tooth NEAR grind*) OR (teeth NEAR grind*))
#14 ((bite* NEAR adjust*) OR (bite* NEAR correct*) OR (bite*
NEAR modif*))
#15 ((occlus* NEAR correct*) OR (occlus* NEAR modif*))
#16 (#10 or #11 or #12 or #13 or #14 or #15)
#17 (#9 and #16)

To identify studies on occlusal treatments, the following

MeSH terms were used: balanced occlusion; occlusal
treatment; occlusal equilibration; occlusal adjustment;
tooth grinding; bite adjustment. To identify RCTs, the
search strategy combined the subject search with the
Cochrane Optimal Search Strategy (see: 15). Search
strategies for other databases were revised appropriately.
Three journals were hand searched: Journal of Oral
Rehabilitation (1974 to April 2002), Journal of Oral and
Maxillofacial Surgery (1982 to April 2002) and Journal of
Craniomandibular Practice (1986 to April 2002).
Additional reports were identified from the reference
lists of retrieved reports and from review articles of
TMD treatments. Unpublished reports or abstracts were
considered from the SIGLE database (April 2002) using
search strategy based on the search strategy presented
above. When necessary, authors were contacted for
relevant original data. Further recommendations were
sought from colleagues on unpublished studies.
The title, abstract and key words of identified studies
were screened independently by both reviewers for
relevance to the systematic review. Studies that
appeared to meet the inclusion criteria were retrieved
as complete articles. Both reviewers independently
extracted data from the included studies to a predesigned data collection form. The data extraction form
considered: bibliographic details, details of the study
setting, characteristics of study population, frequency

and course of the interventions, baseline and outcome

measures, etc. The different requirements and techniques for adjustment and data on psychosocial factors
were recorded as covariates for assessment as possible
sources of heterogeneity. Any adverse reactions were
recorded and described. Uncertainties on data extraction were resolved by discussion between the reviewers. Where necessary, the authors of the original studies
were consulted by mail to obtain more information
about the published study.
Both reviewers independently assessed the quality of
each study according to the guidelines in the Cochrane
Reviewers Handbook. The strengths and weaknesses of
the study design of each included study were analysed.
Disagreements on validity assessment were resolved by
consensus and discussion.
The Cochrane Oral Health Groups statistical guidelines were followed using RevMan 4.1. The studies
were grouped according to types of control and
duration of follow-up.
Cochranes test for heterogeneity was used to assess
discrepancies in the estimates of treatment effects.
A random effects model was used for assessment of
any significant heterogeneity (P < 01) detected. The
source of any statistical heterogeneity was investigated.
A sensitivity analysis was carried out upon different
assumptions such as quality of the studies, whether the
trials were blind or not, missing data and different
statistical approaches. Publication bias was estimated
using the symmetry of funnel plots. The strength
and generalizability of the evidence were carefully
explained.
The proportion of observed and expected agreement
between the reviewers for 45 variables in 17 data
extraction forms was assessed using Cohens kappa. The
test showed a high agreement between the reviewers
(j 088).

Results
The search strategy identified over 660 titles and
abstracts and from this we obtained 23 full reports.
A total of 17 trials were considered eligible according to
the defined criteria for trial design, participants, interventions and outcomes. Of the 17, 11 were excluded
because of attrition bias (three trials), incomparable
duration of intervention and measurement (two trials),
the control groups did not have the condition (two
trials), no valid control group (two trials), invalid

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 287292

289

290

H. KOH & P. G. ROBINSON

treatment group (one trial) and inadequate duration of
measurement (one trial).
Three trials (1618) recruited a total of 92 patients
with symptoms of TMD for treatment and three trials
(13, 19, 20) recruited a total of 299 healthy subjects for
prevention. One prevention trial included young
adults, one trial included adolescents and one study
included children and adolescents.
All included trials had a randomized, parallel group
study design. All six reports provided a clear description
of the type and duration of intervention for both the
test and control group. All but one trial (18) included a
placebo control group. One trial compared adjustment
and reassurance (18). One trial had an additional no
treatment control group (17) besides the test and
placebo groups.
There was variation between the trials in the
outcomes used. Three trials reported both signs and
symptoms of TMD (1618). Two trials (17, 18) reported
data on pain and headache. Two trials presented data
on globus (16, 18). Likewise, there was variation in the
types of measurement used. One trial used a Visual
Analogue Scale for pain and the presence or absence of
headache and globus (18). One trial used the frequency
and intensity of pain and headache (17). One trial used
number of improvements in globus symptoms (16). The
three prevention studies on (13, 19, 20) reported data
on the incidence of TMD.
Additional clinical outcomes reported included range
of mandibular movement (18), disclusion times (17)

Comparison/
outcome

and the presence of an unstable occlusion (16). There

were no data on psychosocial outcomes, costs or quality
of life in any of the trials. There were no reports of
adverse reactions.
Electronic mails were sent to authors of one trial and
data was obtained from one included study (20). The
information supplied was from questionnaires administered pre- and post-treatment regarding symptoms in
subjects who did not request treatment.
No major differences were found in the baseline
characteristics of the groups in terms of the number
randomized, age, gender or the outcomes in Vallon
et al. and Kirveskari et al. (18, 20). It was unclear if
differences in the age and gender existed in one trial
(16), age alone in one trial (19) and gender alone in
two trials (13, 17). There were no major differences in
the other baseline characteristics. The generation of
allocation was adequate in three trials (13, 16, 20),
inadequate in one trial (17) and unclear in two trials
(18, 19).
All the trials were performed by dentists trained in
OA and control. Adjustment for confounders was either
absent or unclear in all trials. The concealment of
allocation was inadequate for one of the six trials (17)
and it was unclear for the remaining five.
Data were analysed on an intention to treat basis in all
except two trials (16, 18). The withdrawals were adequately reported in four trials, unclear in one trial (17). One
trial did not have any withdrawals (18). All but one trial
(17) reported blinding during the outcome assessment.

Number of
studies

Number of
participants

Effect size odds ratio

(fixed) (95% CI)

OA v. placebo
Pain frequency
Pain severity
Headache frequency
Headache severity
Relief of globus

1
1
1
1
1

18
18
18
18
17

050
050
090
090
600

OA v. reassurance
Pain frequency
Headache frequency
Overall symptoms

1
1
1

50
50
50

013 (001258)
140 (045435)
312 (0128040)

OA v. no treatment
Pain frequency
Pain severity
Headache frequency
Headache severity

1
1
1
1

17
17
17
17

010
010
010
010

Table 3. Summary of results

(007385)
(007385)
(013608)
(013608)
(0724984)

(000215)
(000215)
(000215)
(000215)

OA, occlusal adjustment; CI, confidence interval.

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 287292

OCCLUSAL ADJUSTMENT FOR TREATING AND PREVENTING TMD

In patients with TMD, OA did not significantly reduce
the severity or frequency of pain or headache or relieve
globus (all one study each) in comparison with placebo,
reassurance or no treatment (Table 3). In three trials
involving 299 patients without TMD, OA did not
significantly reduce the incidence of symptoms of the
condition (odds ratio 045, 95% confidence interval
015137).
Heterogeneity was assessed for the incidence of
symptoms in the prevention trials. The meta-analysis
shows overlap in the confidence intervals and suggests
that the variation in the results was not because of
chance (P 005). Publication bias was assessed for the
incidence of symptoms in the prevention trials. The
funnel plot is based on three studies and is insufficiently
powerful for any clear indication. The other comparisons had only one trial.
Sensitivity analyses revealed no changes associated
with: changing the inclusion criteria for the duration of
study; using continuous rather than dichotomous
outcomes; using improvement of symptoms rather than
the absence; using random effects models instead of a
fixed effects or vice versa; aggregating the data from the
placebo, reassurance and no treatment groups or
aggregating the data relating to frequency and severity
of pain or headache (all P > 005).

Discussion
There is no evidence that OA treats or prevents TMD.
Data available in the six trials indicate no significant
differences between OA and placebo, reassurance or no
treatment in the treatment or prevention of TMD.
Based on these data OA cannot be recommended in the
treatment and prevention of TMD. These conclusions
will be of interest not only to clinicians but also to
agencies involved in the funding of dental treatment
and remuneration of dentists.
It is important to distinguish between absence of
evidence and evidence of absence. There may not be
evidence of an effect because there are few data
regarding the effectiveness of OA for TMD. The small
number of studies and participants meant that the
confidence intervals were wide. An implication is that
more trials on the effectiveness of OA for TMD
are needed. The inclusion of future trials on
prevention into the current analysis may further
reduce the confidence interval and achieve statistical
significance.

There are concerns of the validity and reliability of

the diagnostic criteria used in the trials. Inaccurate and
inconsistent diagnosis of TMD would cause misleading
reporting of TMD and incomparability of results with
other trials. However, these concerns extend beyond
the data used in the trials to the diagnosis of TMD in
general (21).
Although the sensitivity analyses do not materially
change the results of the review, there are too few
trials, of low quality and with few participants, for the
results to be robust.
There were some limitations of the methods used in
the trials. These limitations should be considered in
their historical context as some of the trials were
conducted some time ago and our understanding of
research methods has improved since then. Guidelines, produced by the CONSORT group, have been
published for reporting of RCTs in the medical literature (22). The use of such guidelines would improve
the quality of trials and reports of the management of
TMD. In particular, future trials of OA should consider
the use of standardized diagnostic criteria and outcome
measures for TMD and providing data on intra- or
extra-examiner variability where appropriate. Reporting the odds ratio, relative risk, relative risk reduction,
absolute risk reduction or weighted mean difference
and associated 95% confidence intervals would allow
a greater understanding of treatment effects and
would facilitate future meta-analyses. Data on psychosocial outcomes, costs and quality of life, and on
any side-effects, especially if they were directly related
to the intervention would also increase our understanding of the potential benefits and harm of this
treatment. Future research should also use samples of
adequate size based on power calculations. The existing trials should be used as the basis of such power
calculations.
In conclusion, this systematic review of RCTs carried
out using the approach of the Cochrane Collaboration
found no evidence that OA treats or prevents temperomandibular disorders. These findings have implications
for dental treatment for patients with TMD and for
researchers in this field.

Acknowledgments
Thanks go to Emma Tavender, Jayne Harrison, Lee
Hooper, Sylvia Bickley and Anne-Marie Glenny of the
Cochrane Oral Health Group for their invaluable help

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 287292

291

292

H. KOH & P. G. ROBINSON

in the editorial process and searching and locating for
the review. Jacob Riis of the Nordic Cochrane Centre
provided Swedish translations. Drs P Kirveskari and
P Alanen (University of Turku) provided additional
information about their trials. We would also like to
thank those who have provided comments and editorial
input into this review.

References
1. Bell WE. Temporomandibular disorders: classification, diagnosis, and management. 3rd edn. Chicago, IL, USA: Year
Book; 1990:289.
2. Dworkin S. Personal and societal impact of orofacial pain. In:
Orofacial Pain and temporomandibular disorders, 1st edn.
Friction JR, Dubner RB, eds. New York, NY, USA: Raven
Press; 1995:15.
3. Rugh JD, Solberg WK. Oral health status in the United States.
Temporomandibular disorders. J Dent Educ. 1985;49:398.
4. Schiffman E, Fricton JR. Epidemiology of TMJ and craniofacial pain. In: Fricton JR, Kroening RJ, Hathaway KM eds. TMJ
and Craniofacial Pain: Diagnosis and Management, 1st edn. St
Louis, MO, USA: IEA Publications; 1988: 1.
5. Puhakka HJ, Kirveskari P. Globus hystericus: globus syndrome? J Laryngol Otol. 1988;102:231.
6. Bell WE. Temporomandibular Disorders. Classification, diagnosis, management, 1st edn. Chicago, IL, USA: Bell WE Year
Book Medical Publishers Inc.; 1986:172.
7. Ramfjord SP. Dysfunctional temporomandibular joint and
muscle pain. J Prosthet Dent. 1961;11:353.
8. Ernst E, White AR. Acupuncture as a treatment for temporomandibular joint dysfunction: a systematic review of randomized trials. Arch Otolaryngol Head Neck Surg. 1999;
125:269.
9. Crider AB, Glaros AG. A meta-analysis of EMG biofeedback
treatment of temporomandibular disorders. J Orofac Pain.
1999;13:29.
10. Harkins S, Linford J, Cohen J, Kramer T, Cueva L. Administration of clonazepam in the treatment of TMD and
associated myofascial pain: a double-blind pilot study. J
Craniomandib Disord. 1991;5:179.
11. Ochs M, La Banc JP, Dolwick MF. The diagnosis and
management of concomitant dentofacial deformity and temporomandibular disorders. Oral Maxillofac Surg Clin N Am.
1990;2:669.

12. Tsukiyama Y, Baba K, Clark GT. An evidence-based assessment of occlusal adjustment as a treatment for temporomandibular disorders. J Prosthet Dent. 2001;86:57.
13. Karjalainen M, Le Bell Y, Jamsa T, Karjalainen S. Prevention
of temporomandibular disorder-related signs and symptoms in
orthodontically treated adolescents. Acta Odontol Scand.
1997;55:319.
14. Forssell H, Kalso E, Koskela P, Vehmanen R, Puukka P,
Alenen P. Occlusal treatments in temporomandibular disorders: a qualitative systematic review of randomized controlled
trials. Pain. 1999;83:549.
15. Shaw WC, Worthington HV, Harrison J, Clarkson JE, Needleman I, Esposito M, Coulthard P, Antczak-Bouckoms A,
Tavender EJ, Bickley SR, Fernandez L. Cochrane Oral Health
Group. In: The Cochrane Library, issue 2. Oxford, UK: Update
Software; 2003.
16. Kirveskari P, Puhakka H. Effect of occlusal adjustment on
globus symptom. J Prosthet Dent. 1985;54:832.
17. Kerstein RB, Chapman R, Klein MA. Comparison of ICAGD to
mock ICAGD for symptom reductions in chronic myofascial
pain dysfunction patients. Cranio. 1997;15:21.
18. Vallon D, Ekberg EC, Nilner M, Kopp S. Short-term effect of
occlusal adjustment on craniomandibular disorders including
headaches. Acta Odontol Scand. 1991;49:89.
19. Kirveskari P, Le Bell Y, Salonen M, Forssell H. Effect of
elimination of occlusal interferences on signs and symptoms
of craniomandibular disorder in young adults. J Oral Rehabil.
1989;16:21.
20. Kirveskari P, Jamsa T, Alanen P. Occlusal adjustment and the
incidence of demand for temporomandibular disorder treatment. J Prosthet Dent. 1998;79:433.
21. Conti PC, de Azevedo LR, de Souza NV, Ferreira FV. Pain
measurement in TMD patients: evaluation of precision and
sensitivity of different scales. J Oral Rehabil. 2001;28:534.
22. Moher D, Schulz KF, Altman DG. The CONSORT statement:
revised recommendation for improving the quality of reports
of parallel group randomised trials. Lancet. 2001;357:1191.
23. Austin DG, Pertes RA. Examination of the TMD Patient. In:
Pertes RA, Gross SG eds. Clinical management of temporomandibular disorders and orofacial pain. 1st edn. IL, USA:
Quintessence Publishing Co; 1995:123.

Correspondence: Dr Holy Koh, Department of Dental Public Health &

Community Dental Education, GKT School of Dentistry, London SE5
9RW, UK.
E-mail: holykoh@dr.com

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 287292