Journal of Oral Rehabilitation

Journal of Oral Rehabilitation 2013 40; 296--318

Review Article
Establishing better biological models to understand
occlusion. I: TM joint anatomic relationships
A. PULLINGER UCLA School of Dentistry, Los Angeles, CA, USA

SUMMARY Belief in and rejection of a relationship specificity 85
7%; 37% versus disc displacement
of occlusion and temporomandibular joint (TMJ) with reduction; and 28
8% versus disc
condyle–fossa position with normal and abnormal displacement without reduction. Significant
function are still contentious issues. Clinical osseous organisational differences between TM
opinions can be strong, but support in most joints with clicking and locking suggest that this is
published data (mostly univariate) is problematic. not necessarily a single disease continuum.
Distribution overlap, low sensitivity and specificity However, a subset of joints with clicking contained
are a common basis to reject any useful prediction characteristics of joints with locking that might
value. Notwithstanding, a relationship of form contribute to symptom progression versus
with function is a basic tenet of biology. These are resistance. Moderately strong models confirm there
multifactor problems, but the questions mostly is a relationship between TMJ osseous organisation
have not been analysed as such. This review moves and function, but it should not be overstated.
the question forward by focusing on TM joint More than one model of normals and of TM
anatomic organisation as the multifactor system it derangement organisation is revealed. The
is expected to be in a closed system like a synovial implications to clinical decision-making are
joint. Multifactor analysis allows the data to speak discussed.
for itself and reduces bias. Classification tree KEYWORDS: temporomandibular joint, anatomic
analysis revealed useful prediction values and relationships, multifactor, models, prediction,
usable clinical models which are illustrated, function, derangement, normal, abnormal
backed up by stepwise logistic regression.
Explained variance, R2, predicting normals from Accepted for publication 3 January 2013
pooled TMJ patients was 32
6%, sensitivity 67
9%,

cally expected but does not have much support until

Introduction
examined in multifactor models (1–3). However, the
The purpose of this paper and the CORE China 2011 strength should not be overstated. A subsequent
conference, of which this was the opening presenta- paper by the author also presented at CORE China
tion, is to revitalise the subject of ‘occlusion’ which 2011 examines for multivariate relationships between
seems to be stalling, with most questions unanswered, dental occlusal variation and temporomandibular dis-
despite 100 years of study and experience. Past key- orders (4). A grand multifactor model integrating nor-
stone literature is highlighted but a comprehensive mal–abnormal TM joint relationships with dental
review of all the past literature is not attempted. This occlusion is a future goal but will be mathematically
paper reviews for a prediction relationship between difficult to achieve owing to the number of entry
temporomandibular (TM) joint osseous organisation variables and the fact that all multifactor analyses and
and normal–abnormal joint function, which is clini- models are relative to the entry variables used. The

© 2013 Blackwell Publishing Ltd doi: 10.1111/joor.12032

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 297

objective of this paper is to add biological perspectives the slower changes in most orthodontics. Rapid adap-
to previously defined technique models. Part of this is tive capacity is needed to adjust to orthotic appliances;
an effort to build multifactor multivariate models (1– the rapid changes in an occlusal equilibration (e.g. to
3, 5–10) to represent and explain what we all already centric relation occlusion), or completing an occlusal
refer to as a multifactor disorder, yet rarely analyse as reconstruction over only a few weeks, including the
such. adaptation demand of conversion to a more vertical
More than one model for normal, and more than pattern of mastication (15, 16); or simply placing resto-
one model for disease, is revealed (1–3), confounding rations. However, adaptation is not an excuse for ‘any-
the validity of traditional univariate analysis including thing goes’ in dentistry: instead, our treatment goals
its ability to reject a relationship. However, it will be must be to minimise the adaptation demand (6) we
shown that TM joint osseous organisation (1–3) at impose on our patients.
best explains only part of the differences between Current successful adaptation probably depends on
normals and disease groups. The current models can age (12), the size and rate of change which may com-
be criticised because they are based on disease preva- press the imposed physiological adaptation demand,
lence and relative risk of disease, whereas what we irritability of the neuromuscular system (17, 18), pre-
really need clinically are models of future risk built sence of normal versus unstable or pathological joints,
on incidence of disease. Meanwhile, the author sug- and whether a single and stable end point of jaw clo-
gests concentrating on occlusal and orthopaedic nor- sure is established occlusally to carefully programme
mals models typifying current normals from defined the pattern of jaw movements (15) and the neuromus-
subdiagnostic groups of patients with temporomandib- cular engram (19). A precise occlusal position may be
ular disorders (TMD) as the best starting point. more important than the jaw position (20). Risk avoid-
ance dentistry might include, where appropriate, try-
ing to practise Centric Occlusion Dentistry to avoid
What is normal?
introducing an extra orthopaedic component in a
Normal for the last 1 million years, until the advent patient with TM joint dysfunction, or, if making large
of agriculture, was loss of all occlusal anatomy and changes make them slowly to permit adaptation and
incisal edges, plus extensive proximal wear and mesial learning. The conformative approach to dental occlusal
drift developing a Class III style occlusion by age 12– treatment (21) elects to maintain the existing intercu-
15 (11). The common features included development spal occlusion position (presuming in the case in ques-
of a reverse curve of Sp
ee and a very flat occlusal tion it is associated with normal function) because it is
plane, a reverse curve of Wilson, more upright posi- thereby said to be less likely to induce problems for
tioning of the anterior teeth, possible active eruption the tooth, the periodontium, the muscles, the tempo-
of teeth vertically to compensate for wear (currently, romandibular joints (TMJs) and the patient.
we only considered passive eruption until teeth out of
contact re-establish contact) and significant adaptive
TM joint adaptation
changes in the TM joints (11) due to the progressive
forward condyle positioning during childhood and McNamara (12) showed an adaptive increase in the
adolescent years (11, 12). Most of the TM joint and fibrocartilage layer of the posterior–superior part of
occlusal adaptation must therefore have occurred the condyles and post-glenoid process with continu-
slowly during development. Some adaptive changes ous protrusive jaw position function in juvenile grow-
would have continued in adulthood. ing rhesus monkeys. It was hypothesised this was a
We must have the exact same capacity, and without cartilage response to traction/pull and to electrophysi-
it, even the most precise dentistry would not work. This ological changes in lateral pterygoid muscle function.
includes, for example, the major neuromuscular, It was followed by deposition of new bone in the con-
orthopaedic and loading changes inherent in orthogna- dyle until structural balance was restored, after which
thic surgery, and major vertical dimension of occlusion the proliferation of cartilage is reduced and there is a
changes (13), all requiring muscle and histological return to the original pattern of muscle activity. Less
adaptation (14); also orthopaedic orthodontics, such as response was seen in adult animals. Revisiting the
functional appliances and crossbite corrections, versus Herbst appliance concept designed to anteriorise a

© 2013 Blackwell Publishing Ltd

298 A. PULLINGER

Class II, one occlusion, expecting condyle remodelling in a logistic regression study with disc displacement
to compensate, may therefore have a place in theory with reduction (DDw/R) as the dependent variable,
despite reports of mixed or neutral outcome (22, 23), and posterior crossbite, gender, age and previous or
but should probably be used only in the young (12). current orthodontic treatment as the independent
variables, none of which were significant, nor were
significant if limited to those without an orthodontic
Functional crossbites are nature’s
history (38). This was de facto a study of a single
continuing experimental orthopaedic
occlusal variable combined with demographic and
model
treatment co-factors. Unilateral crossbite was found in
A classical Angle’s Class II, division 1 unilateral func- 12
2%, but only 4
1% of the total sample had TMJ
tional crossbite induces a geometric shift in the jaw clicking. Some inherent problems with logistic regres-
and condyle positions, displacing the condyle on the sion models exist. Only a single model is generated,
crossbite side posteriorly and superiorly (24) accompa- and in this case it is a problem that in most cases
nied by asymmetric jaw posture muscle activity (25– adapts successfully skeletally and dentally rather than
27). A functional crossbite usually adapts successfully at the expense of the disc, or at least as stated in this
and stabilises as a skeletal crossbite by young adult- article, had mostly adapted successfully until young
hood by dental alveolar, skeletal and TM joint adapta- adolescence. Other studies have compared normals to
tion changes (28–33) and the jaw can then no longer patients seeking treatment in a TMD/oro-facial pain
be manipulated back into a more non-crossbite posi- centre, implying a greater symptom problem level.
tion. Unilateral crossbite in other Angle’s classes Stratifying the sample (38) a priori into good and poor
should not be excluded but was not similarly studied adaptation responders and examining many more
in numbers. An imposed physiological adaptation variables in broader multifactor logistic regression
demand on the stomatognathic system would require analyses would be more interesting. Construction of
a functional displacement component so anterior prediction models is our goal. The same paper
teeth in crossbite would probably not qualify. reported that orthodontically treated subjects did not
Although there is not complete agreement (34), have more TM joint disc displacement clicking, but
unilateral functional crossbites have been associated this also shows that it did not reduce the prevalence.
with disc displacement derangements (35). In con- Owing to the published small measurable relative
trast, bilateral crossbites seem to have less orthopaedic risk, correction of a unilateral functional crossbite is
consequences and were not associated with TMD still recommended in a child to reduce the expected
symptoms in children and teenagers (36). However, imposed physiological adaptation demand by return-
the predictive value of unilateral functional crossbite ing condyle–fossa positions to more bilateral symme-
for disc displacements may be quite low (37) due to try and more centricity early (24, 28). However,
successful skeletal adaptation in most cases rather ‘prophylactically’ normalising a similar crossbite in an
than adaptation at the expense of the articular disc. adult is too late and redundant if a TM disc displace-
Notably, a prospective MRI study revealed that only ment derangement has already occurred, and ‘correct-
1/15 children with unilateral crossbite occlusions had ing’ a unilateral crossbite in an asymptomatic adult to
a disc derangement (locking) (34). In more complex prevent future TMD problems is not supportable if
multifactorial stepwise linear regression models (5), TMJ skeletal remodelling has successfully stabilised
unilateral posterior crossbite contributed as a co-factor the jaw system (32).
with additional occlusal variables differentiating disc Other associations with unilateral functional cross-
displacement with and without reduction, jaw myal- bites have been reported, reflecting the physiological
gia-only patients and osteoarthrosis with an earlier adaptation that a functional shift can impose, includ-
history of derangements, from normals, but this is ing condyle subluxation laxity (39), and more frequent
within the context that the total explained variance deviation-in-form remodelling in all components of
of the models was still only in the 4
8–13
7% (R2) the TMJ in young adult autopsy specimens (40), in
range for the derangements (6). which more articular changes were also reported in
Unilateral posterior crossbite was not associated anterior crossbite cases. Several studies relate
with TMJ clicking in a large young adolescent sample right–left-side differences in condyle and fossa shape,

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 299

size and position (41–44) to mandibular-occlusal and functional registration techniques using an
asymmetry, not presenting as a crossbite. In an imag- anterior deprogrammer and masseter contraction (53).
ing study of young people comparing symmetrical ver- The differences in the alignment of condyle osseous
sus asymmetrical Class III occlusions, more TMJ trabecular pattern between Angles Class I, Class II and
symptoms (54
5% vs. 38
6%) and disc anterior dis- Angles Class II with an open bite suggested in a very
placement (56
8% vs. 18
2%) were found in the small sample skull study (54) might be markers for
asymmetric group (45). differences in joint loading presumably from different
patterns of jaw function or use. Adequate examina-
tion will doubtless also require multifactorial studies
Fossa and eminence development
including, for example, Angle’s class broken down
There are many ideas about the possible influences of and represented as occlusion variables, anterior attri-
the geometry of the fossa and eminence to normal tion/bruxism scores and habits, condyle–fossa posi-
and abnormal TM joint function with particular tradi- tion, and craniofacial profile/angles affecting jaw
tional interest in the articular eminence slope. The muscle power and direction (55).
articular eminence is developmentally dynamic and The articular disc is considered an important part of
adapts rapidly until the completion of deciduous den- TM joint stress distribution, and disc thickness is a
tition. The fossa is approximately flat in newborns, critical factor determining the peak stresses measured
and its loading is rather vertical, with the articular under the disc (56). Static loading becomes traction
eminence developing probably by loading stimulation loading that increases after compressive strain at the
as the pattern of the primary dentition is established start of condyle translation movement and is velocity
(46), attaining about 45% of its adult angle value by dependent during movement. It is suggested that trac-
age 2, 70–72% by age 10 and 90–94% by age 20 tional forces may contribute to mechanical fatigue
(47). Conversely, the slope of the articular eminence and degeneration of the disc (57). TM joint loads
and the height of the fossa were reduced in a sample were calculated 9
5–69% higher in individuals with
of patients who were edentulous for more than disc displacement versus with normal disc position
3 years (48), as well as in an edentulous–age interac- (58), which might contribute to mechanical failure of
tion (49). The fossa is concluded to be a remodelling the disc and potential degenerative joint disease.
unit and changes slowly when function is dramati- Risk for mechanical failure of the TMJ disc could
cally altered as in the case of edentulism (50). reflect differences in TMJ loading. However, as dis-
Although included as an entry variable, eminence cussed later, only a subset of individuals with TMJ
slope was not retained in the authors’ anatomic mul- clicking progress to disc displacement without reduc-
tifactor models for the TM joint normal-derangement tion (DDw/oR) and only a subset of these develop
function (1–3) but may have been subsumed in other osteoarthrosis. Conversely, many individuals with
retained variables of eminence radius, length and TMJ osteoarthrosis deny any prior history of TMJ
height, fossa depth and width. internal derangement (59).

TM joint loading Is there a condyle–fossa position

associated with normal function?
Synovial joints by definition are loaded joints. There-
fore, the TMJ is also loaded. The mapping of the The optimum position of the TMJ condyle–fossa posi-
fibrocartilage layer extends only from the crest of the tion is a concept with many clinical opinions (60–63).
roof of the fossa, down and across the articular emi- The two main clinical dentistry concepts are geomet-
nence (51), and is therefore a proxy for the location ric centricity, or other geometry, and belief by many
of joint loading signifying the predominant superior– that centric relation is the ideal functional maxillo-
anterior vector of loading in the TM joint. This mandibular relationship defined as that in which the
matches the net vector of the mandibular elevator condyles articulate with the thinnest avascular por-
muscles, and matches the accepted definition of cen- tion of the disc with the disc–condyle complex in the
tric relation (52), plus seems compatible with the anterior–superior position against the slope of the
bimanual manipulation technique for CR registration articular eminence (52). The author prefers instead

© 2013 Blackwell Publishing Ltd

300 A. PULLINGER

the medical term and concept of the ‘Closed-Packed derangement, and disc displacement was also associ-
Position’ (64) = the joint position where there is maxi- ated with changes of condyle deformity and other
mum congruity (or seating) of the articular surfaces. degenerative changes (71). Condyle position measured
What happens to ‘centric relation’ if there is disc or on the main symptom side was reported as more fre-
condyle pathology or even no disc? Is this to be quently posterior in a disc derangement group 52
5%,
termed ‘pathological centric relation’, sometimes concentric 40% and anterior 7
5%, compared to 15
8%
referred to as an ‘adapted centric relation’ or an ‘adapted posterior, 78
9% concentric and 5
3% anterior in a
centric posture’ (65). Notwithstanding, a clinical or myalgia-only group (72), and a more equal broad dis-
manipulated centric relation can still be a useful refer- tribution of positions in a TMJ osteoarthrosis group.
ence and transfer position, but it does not have to be Therefore, although of clinical and biological interest,
the treatment position. posterior position cannot be considered diagnostic of
A published literature review (66) concluded that TMJ derangements (73, 74). Consequently, TM diagno-
normal condyle–fossa position at the closed intercu- sis and treatment of TMJ disorders should not be based
spal position is approximately concentric, but there is solely on radiographic position of the condyle (75).
a lack of agreement because both normals and disc Additional studies report more posterior positions in
derangements have a wide distribution range (67) mild to moderate disc derangements, but interestingly
and thus would fail in univariate prediction value. were reported as more concentric when the derange-
Condyle–fossa positions in asymptomatic normals ment became severe (76, 77). This probably suggests
with no history of orthodontic or extensive restorative some dichotomy or subgroups in disc displacements
treatment were reported to be predominantly with severity, and with and without reduction.
concentric, 53%, but included 23% posterior and
24% anteriorly located condyles (68). Some right–left
Posterior ‘displacement’ of the condyle
asymmetry can be normal (69). Therefore, non-
and mandible versus ‘posterior condyle
concentric condyle positions can be compatible with
position’
normal function although a prospective outcome
study has not been done. Interestingly, in the tail dis- Dentists frequently use the presumptive causation
tributions in the study of asymptomatic normals, term posterior condyle displacement instead of the more
there was a shift to more anterior positions in men neutral term posterior condyle position. The following
and more posterior positions in women (68). If pos- aetiologies have been suggested.
terior position does contribute any risk for disc dis-
placement, could this partially contribute to the
Does disc displacement cause a ‘displacement’ of the
greater frequency of TMJ clicking in females?
condyles?
Although reciprocal clicking occurs in both sexes,
closed lock as a severe derangement is very rare in This would imply that anterior disc displacement is an
men in the absence of trauma or osteoarthrosis. This initiating event, inducing a reciprocal posterior dis-
is why the authors’ TM joint derangement study sam- placement of the condyle and mandible. However,
ples are limited to females in the multifactor studies. this seems highly unlikely because this would require
over-riding the effect of the intercuspal position,
which usually defines the jaw and condyle–fossa posi-
Is non-concentric condyle position
tion. Some slow accommodation to a slightly superior
associated with, or diagnostic for, TM
position for the condyle could be more realistically
joint derangements?
hypothesised as the thicker posterior band of the
Condyle–fossa position in TMJ derangements is often fibrous part of the articular disc adjusts or dislocates
described as skewed towards posterior positions, and anteriorly. A difference in condyle position is reported
some authors have attributed some diagnostic value to in one study between normals and DDw/R cases, but
that observation (70). In a large MRI study of patients not versus DDw/oR, which is a more complete disc
with disc internal derangements, condyle positions displacement (77). This would seem to be contradic-
moved from a mean centric position to an eccentric tory for a disc anterior displacement aetiology for con-
dorsal position with increasing severity of the internal dyle posterior displacement.

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 301

The conclusion of one CBCT and MRI study of TM In contrast, the average condyle positions of Class II
joints spaces and disc position in 12- to 20-year-olds symptom-free individuals with deep bite in another
post-orthodontic treatment was that disc displacement study were concentric and did not differ from controls
can alter the condyle position in the fossa and its joint (85). No specific correlation of condyle position with
space dimensions, depending on the direction and deep bite and steepness was shown (86), or to overbite
extent of the disc displacement (78). The mean pos- or overjet, or when factored together as a proxy for
terior joint space was significantly less in partial and steepness (87). No differences at all were shown in the
complete DDw/R than optimal normal measurements, cumulative distribution of continuous measures of
but with no differences between the two disease overbite and overjet between DDw/R or without
states. However, superior joint space was smaller in reduction and normals (59). Therefore, the authors
complete than in partial displacement. The mean cen- recommend using continuous measures of overbite
trally measured joint space was also smaller in lateral and overjet (as for most measures) wherever possible
and in medial disc displacement than optimal nor- to permit the data to speak for itself rather than the
mals, and the joint space was larger laterally in lateral potential bias of cut scores defining deep bite. These
disc displacement and larger medially in medial disc data therefore question the concept that a steep ante-
displacement. The current author believes that some rior occlusion displaces the jaw posteriorly, or at least
minor adjustment of the condyle to a slightly more not in the naturally occurring occlusion.
superior position is possible following displacement of
the thicker posterior band of the disc anteriorly at the
Can the occlusion displace the mandible and condyles
closed jaw position, and very slight associated vertical
posteriorly (in treatment occlusions)?
adjustment of posterior teeth. However, a change in
mandibular antero-posterior and lateral-medial posi- However, conceptually, a rapidly induced mandibular
tion seems more complicated. The cross-sectional retrusion shift could carry a greater risk for posterior
study design (78) cannot differentiate whether this is displacement of the mandible, for example, by a rapid
instead an adjustment of disc position secondary to incisal orthodontic pull-back, prosthodontic recon-
the pre-existing dimensions and ratios of the joint struction to a more vertical anterior guidance and
spaces, versus whether disc displacement is the equilibrating out a large CR-CO (RCP-ICP) occlusal
primary aetiology? The post-orthodontic treatment slide. The influence, timing, speed and forces in
sample also raises possibility of an orthopaedic–ortho- orthodontics is a potential question (88). Fortunately,
dontic treatment effect as a confounding factor. most orthodontics is slow and allows successful
adaptation in most cases. On the other hand, most
orthodontic patients are in the TMJ derangement-
Can the occlusion displace the mandible and condyles
susceptible young age group versus prosthodontics,
posteriorly (in the natural occlusion)?
which is usually in an older age demographics. Not-
There is a common belief that a deep bite, a Class II withstanding, condyle position is reported as remain-
division 2 occlusion, and retro-inclined incisors cause a ing in a stable and an overall slight anterior position
posterior displacement of the mandible and condyles despite orthodontic treatment, with only 9/222 joints
leading to TMJ clicking or joint pain (79–81). Some ending in a posterior position (89, 90). Condyle posi-
correlation was shown in one small sample study of tion was also reported to be unrelated to extraction or
deep bite with sleep and daytime clenching, headache, orthodontic treatment, bite depth, interincisal angle
bite discomfort, jaw stiffness, muscle problems, disc dis- and upper incisor inclination in those studies.
placement and other joint disorders, as well as psycho- A 15- to 18-year follow-up of orthodontically treated
logical factors (82). Deep bite with TMJ clicking and adolescents (91) supports the opinion that orthodontic
bruxism were significant but overall weak predictors of treatment in childhood did not seem to increase risk
TMD signs and symptoms in young adults in a 20-year for signs or symptoms of TMD later and carried only
follow-up logistic regression study (83). More posterior a 1% incidence per year of TMD issues requiring
condyle position in Class II, 2 is supported in another treatment. Nevertheless, the corollary to group studies
study (84), but as is usual, the sample was small reflect- reporting that orthodontics, which is designed to
ing the low population prevalence of this Angle’s class. normalise or idealise the occlusion, does not cause

© 2013 Blackwell Publishing Ltd

302 A. PULLINGER

TMD (92) is that it does not prevent TMD either. (97). Histological changes were induced in the condy-
More importantly, orthodontics itself is an exemplar lar cartilage that appeared to affect growth resulting
of a multifactorial problem both in patient typing and in a retrognathic pattern.
in treatment methodologies, some with potential
orthopaedic features, requiring multifactor modelling
Does TM joint disc displacement cause skeletal alteration?
for analysis, and examination of subsets.
Prosthodontics can suddenly impose a more vertical Differences in cephalometric measurements of skele-
envelope of motion on a previous horizontal chewing tal morphology and symmetry were described
habit (15), which can upset a few patients especially if between asymptomatic normals and patients with
they have an irritable stomatognathic system and or a disc derangements (98) but not when compared to
continuing strong protrusive bruxism or forward jaw TMD patients (presumably myofascial pain patients)
posturing habit. Individuals with a previous habitually with bilaterally normal asymptomatic TMJs (99).
wide lateral chewing stroke tend to touch the palatal However, no significant differences in cephalometri-
surface on an imposed steeper guidance before termi- cally imaged dental pattern were described. The
nating in their intercuspal position (16). Nevertheless, authors (98, 99) concluded that disc displacements
while it is clinically self-evident that most will adapt may affect the skeletal morphology and especially
normally, a few feel vertically trapped, some may brux the mandibular position and rotation and might
down the increased incline, the teeth may move or cause skeletal alteration. The current author suggests
exhibit fremitus, or muscle soreness might be activated. that skeletal variation may instead precede and affect
To reduce risk, providing a longer adaptation period the patterns of TM joint function and loading,
(best provided in the provisional crown phase, also per- thereby contributing risk for disc instability versus
mitting changes to be made and then copied in the final orthofunction. Certainly differences in jaw muscle
restoration) is advised. vectors and forces are demonstrated with skeletal
variation (55). In contrast, a secondary malocclusion
and skeletal change secondary to bilateral TM joint
Is disc displacement part of an arthritis process?
osteoarthrosis (100) does seem likely. Adding care-
In another TMJ imaging study, it was suggested the fully measured craniofacial skeletal variables, joint
TMJ disc derangements were less influenced by TMJ loading and muscle vectors to TM joint-fossa organi-
morphology, disc and lateral pterygoid factors or sation would be wise for testing in a broader multi-
occlusion (93). Instead, the only common finding and factor orthopaedic model.
conclusion was that impairment of joint lubrication
plus local and systemic factors ought to be investi-
Does condyle–fossa position define the
gated (94). It has been suggested that disc derange-
disc shape?
ment is a part or subset of an osteoarthritis process
and that more weight ought to be attributed to loss of Conceptually, a more stable biconcave disc shape is
fibrocartilage and inflammation (95). expected with a concentric or slightly anterior con-
dyle–fossa position, which defines the available joint
space shape (Fig. 1). Conceptually, a wedge-shaped
Does disc change lead to altered condyle shape and size?
disc that is thinner posteriorly is considered geometri-
One study (96) presumed that the disc architecture is cally to have less stable passive biomechanics and to
the primary aetiology leading to change in shape and be at greater risk to adjust or displace anteriorly over
smaller size of the condyles. However, it would seem time. A wedge-shaped disc shape seems more likely
as likely that the disc is accommodating to alteration with a posterior condyle position if the available supe-
in joint architecture or they are co-adaptive. rior joint space is then rendered smaller than the
anterior joint space. However, by itself this would
seem to be insufficient to induce disc displacement
Does change in disc physiology affect mandibular growth?
because a number of normals have posterior condyle
Surgically induced bilateral TM joint DDw/oR in juve- positions, and many joints with disc displacement are
nile or adolescent rabbits impaired mandibular growth concentric.

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 303

(a) (b)

Concentric Posterior

(c) (d)

Fig. 1. Showing (a) temporoman-

dibular joint lateral mid-section
tomogram with concentric condyle–
fossa position; (b) a posterior condyle
Stable passive biomechanics Wedge shape potentially unstable
position; (c) conceptual accommodation passive biomechanics
of a stable disc shape in (a); (d) a
potentially more unstable wedge- (e) (f)
shaped disc in the wedge-shaped
disc space in (b); (e) normal tempo-
romandibular joint disc histology, shape,
attachments and lateral pterygoid
muscle insertions; (f) a pathologic TM
joint with posterior condyle position,
unfavourable disc shape and forward
position, and osseous changes.

It is reported in one pre-orthodontic treatment sam- and matches up with the mediotrusion side move-
ple (84) that when condyles were anterior or concen- ment of the condyle in a lateral jaw movement, and
tric, the discs were in a normal position and had to the angulation of the lateral pterygoid muscles
normal biconcave shape. When condyles in normals (107).
were located posteriorly, most articular discs still kept In contrast, the earlier idea that anterior disc dis-
a normal biconcave shape and maintained a normal placement was caused by spasm in the lateral ptery-
or slightly anterior position. However, some of the goid muscle (108) is still heard but may have
discs were significantly more anterior and had abnor- doubtful anatomic support because only a few fibres
mal shape and therefore constitute a subgroup that of the superior belly of the lateral pterygoid insert
should be followed prospectively. Degrees of anteri- separately into the articular disc (109), even though
orly located discs in normals is described on MRIs in two separate bodies of the lateral pterygoid can be
20–34% (101–103) and is considered a variant of nor- identified at their origin and do function differently
mal in those cross-sectional studies. Disc positions up (110). A different model with the disc attaching onto
to +15 degrees on medial tomograms and +30 degrees the superior border of the superior lateral pterygoid
on lateral tomograms have been regarded as normal muscle like a snails foot (111) is a better representa-
variations (104). Therefore, a posterior condyle posi- tion, but with the majority of the lateral pterygoid
tion per se is not diagnostic of anterior disc displace- muscle fibres actually inserting into the fovea of the
ment. condylar process (109) (Fig. 1e).
The importance of passive biomechanics is obvious
in videos of TMJ autopsy specimens (105) in which
Angle’s class versus Condyle–fossa
normal and abnormal disc function matches our clini-
position and jaw dysfunction
cal and MRI experience, when the examiner moves
the mandible. Although the literature continually There is no obvious relationship of Angle’s class with
refers to anterior disc displacement, in reality an jaw dysfunction (59). Perhaps, Angle’s classification is
antero-medial displacement is more common (106) too crude a parameter. However, it does seem pre-

© 2013 Blackwell Publishing Ltd

304 A. PULLINGER

sumptive to call Class I ‘normal’, and Classes II and III dyle position anteriorly in response to an early estab-
‘malocclusion’/abnormal, when Class III edge-to-edge lished forward jaw posturing trait for incision and
occlusion has previously predominated the adult speaking during development, perhaps also normalis-
human dentition for >1 million years (11). Phyloge- ing the facial profile, and possibly improving the oro-
netically, if Classes II and III were ‘malocclusions’, pharyngeal space and airway. In the asymptomatic
thereby detrimental to masticatory function and sur- normal young adult sample (87), there was no rela-
vival, they would have long since disappeared from tionship of condyle position or Angle’s class with
our gene pool. The author strongly recommends using RCP-ICP (CR-CO) occlusal slide but the range of slide
the term occlusal variation instead of malocclusion (112), in the sample was not large and skews to a low end
which should be reserved for occlusal organisation range. Asymmetric bilateral condyle position occurred
proven detrimental to dental hard and periodontal tis- significantly more frequently in the subset of normals
sues and if actually causative of jaw pain or dysfunc- with asymmetric occlusal slide and least in normals
tion. with no slide (87). Possible relationship of condyle
It does not seem to be useful to include the interac- position with Class II, 2 is discussed above. Condyle
tion term Angle’s class as an entry variable in multifac- position in Class IIIs is described as being concentric
tor studies because it can probably be better or more anteriorly positioned (84–87, 114), suggesting
represented by its separate variables. Only two associ- a more forward jaw function matrix. Some shape dif-
ation examples were found and were relatable to ferences in condylar and eminence inclination and
Class II: firstly in a sample of patients with myofascial fossa width and depth are described between Angle’s
pain  limited jaw opening and arthralgia (113) and Class II and Class III (115).
a slight increased odds ratio prediction of DDw/oR
with increasing distal occlusion as a co-variable (5).
Evaluation methods
However, this needed to be at a 9-mm distal posi-
tion to reach a 2:1 odds ratio, whereas the largest
Subjective observation
measurement in the actual sample was 5
5 mm (5).
In a young adult autopsy histological study, more Subjective observation of condyle–fossa position may
TMJ condyle and fossa deviation-in-form and disc dis- have some clinical utility and in one study had an in-
placement were found in cases with abnormal/larger terobserver agreement of 69–79%, and intra-observer
overjet, and Angle’s Classes II and III dentitions com- agreement of 81–90% (116). The range 12% on the
bined with age were significantly associated with quantitatively measured joint space ratio method
more deviation in form. Also more Class II dentitions ((posterior–anterior joint space/posterior+anterior
had histological evidence of TMJ remodelling changes joint space)%), frequently used to express condyle
(40). However, remodelling does not have to mean position (117), corresponded to subjective judgments
pathology. of concentricity on TMJ tomograms.
Although a wide distribution of condyle positions is
reported for all Classes (84), there is some repetition
Panoramic imaging
that condyle–fossa positions at the intercuspal position
differ descriptively between the different Angles clas- The author considers TMJ standard panoramic imag-
ses. Although not supported in all studies (86), Class I ing to be usable as a screening view but is not consid-
asymptomatic normals with no history of orthodontics ered sufficient for TM joint diagnostic purposes (118).
or much restorative treatment were significantly more
concentric with 25% posterior, 57% concentric and
Transcranial radiographs
18% anterior condyle positions. This was compared to
a shift towards more anterior positions in Class II, 1 Full concordance for condyle position in subjective
with 10% posterior, 40% concentric and 50% ante- evaluation between transcranial views and linear to-
rior positions (87) as also reported by others (107). It mograms was found in only 60%, and the transcra-
may be hypothesised in Class II, 1 that this may nials tended to exaggerate non-concentricity in 30%
reflect a functional matrix that established the (119).
intercuspal occlusion and thereby the matching con-

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 305

Cone beam CT imaging sections (120). The number of data points needed to
represent the fossa and the disc tissues in three plane
It is now considered the clinical standard of care and
sections and especially in true 3-D will seriously tax
a new research norm. The author believes the indica-
the capability of current multifactor analysis pro-
tions for CT tomograms (optimising hard tissues and
grammes. Perhaps, there are ways to more simply
pathology) and MRIs (optimising detail in soft tissue)
represent 3-dimensionality through different patterns
are not interchangeable, so no imaging is considered
of agreement or differences between sagittal, lateral
universal clinically and that each has their indica-
and medial sections. Normative joint space measure-
tions.
ment values in the coronal or axial plane have now
been provided for normals (121), which could be eas-
Univariate measurement of CT imaging ily run with logistic regression analyses of sagittal sec-
tion data to develop three dimensionality. It has been
Despite a trait for more posterior position in the disc
pointed out that some patterns of progressive increase
derangement diagnoses, there is a diagnostic impasse
or decrease in joint space from medial to lateral sec-
owing to the great overlap with the wide distribution
tions might actually represent asymmetric geometric
in normals (1) (Table 1). Therefore, condyle position
displacement of the mandible (68, 122), especially if
per se is not diagnostic and would fail any useful pre-
continued contralaterally.
diction values.

Multifactor studies of TMJ osseous

Multivariate measurement of CT imaging (to date mostly
organisation
represented by centre condyle sagittal sections)
The author frequently points out that Mother Nature
The authors’ therefore extended the examination
is not interested in ‘which’ answer is correct in a mul-
through multifactor analyses, because it is illogical to
tiple choice question: instead, the answer is ‘all the
proffer that there is no relationship between form and
above’ but ‘it depends’. Hence, there is the contempo-
function, and especially in a closed system like the
rary interest and need for multifactorial analyses and
TMJ synovial joint. The TM joint has to be examined
prospective studies when presented with soft but
as a whole, starting here with the organisation of the
sometimes intriguing univariate data.
osseous architecture on tomographic radiographs.
The multifactor models presented used 14 linear
and angular measurements with eight ratios of con-
3-D measurements of joint spaces and anatomy dyle and fossa shape, size and joint space measure-
ment as entry variables (1–3). The analyses suggest
Three-dimensional measurements and assessment
that there are significant interactions between fossa
have been undertaken although are usually limited to
size and shape with condyle positions, including joint
assembled representative lateral, central and medial
space ratios as a proxy for the available disc shape
space that can differentiate between normals and clas-
ses of disc derangement and between derangement
Table 1. Condyle–fossa position index* univariate data
classes. Hence, they may be players in determining
[(P A)/(P + A)% joint space] (1)
disc stability–instability orthopedically.
Mean Mean  1 s.d. Range

Asymptomatic 4
34 23
77 to 15
09 66
83 to 47
26 Differentiation of normals (multifactor analyses)
normals
A classification tree analysis (1) appears like a genea-
Disc displacement 13
76 37
9 to 10
38 55
13 to 63
45
with reduction logical tree (Fig. 2) that classified and differentiated
Disc displacement 13
30 38
51 to 11
91 61
90 to 54
14 three patterns of normals from four subgroups of
without pooled TMJ disc derangement patients. The complete
reduction model approaches clinically useful specificity (123) at
*Index: 0 = complete concentricity; ve = more posterior; 85
7% differentiating normals, which is the group
+ve = more anterior. purpose, and moderately good sensitivity (67
9%)

© 2013 Blackwell Publishing Ltd

306 A. PULLINGER

Concentric to very Concentric to very Concentric to very Concentric to

posterior <–5·08> anterior posterior <–5·08> very anterior
Condyle position index
Condyle position index
Average to narrower Average to wider/
/and or deeper <2·74> and or shallower Average to narrower <17·86> Much wider
Fossa width/fossa depth
Fossa width
1: DD 2: DD
ASx = 1 Concentric <7·04> Slightly to very
ASx = 1
DD = 57 anterior
Much deeper <2·43> Average Condyle position index DD = 19
Fossa width/fossa depth

5: ASx
3: DD 4: ASx ASx = 10 Very shallow <5·72> Average
ASx = 1 ASx = 4 DD = 10 Fossa Depth
DD = 22 DD = 28

6: DD 7: ASx
ASx = 0 ASx = 4
DD = 12 DD = 14

Fig. 2. Classification tree analysis of the osseous anatomic organisation of TM joints on tomograms, differentiating asymptomatic nor-
mals from pooled temporomandibular joint disc displacement cases.

differentiating and predicting patients. The guidelines

Normals 1.
for TMD recommend sensitivity >75% and specificity (vs. pooled
>90% (123) to avoid over diagnosis and erroneous or derangements)
over treatment in TMD. As a non-life-threatening 47·6%
–5% to +7% P-A %
pathology we can accept some false negatives, while P+A
maximising the correct identification of normals.
The amount of difference (explained variance, R2)
Normals 2.
in the model between normals and derangement
patients explained by the orthopaedic variables was & average < –5%
posterior disc 19·1%
high (32
6%). Conversely, this should not be over
space if condyle
read clinically because it also means that 67
4% of was very
posterior
the variance is unaccounted for. Therefore, additional
tissue factors need researching to include in a larger Normals 3.
model, because they might be as or more important.
Further differentiation and complexity may be possi- > +7%
19·1%
ble in the current sets with larger sample sizes, but it
is not expected to see the explained variance rise very
much more. This is cross-sectional prevalence-based
data, so care should be taken before attributing Fig. 3. Clinical illustration of the three osseous anatomic organ-
isations of normals (models 1–3) shown in Fig. 2.
aetiology.
The usefulness of the classification tree analysis is
that the models can be represented in real clinical characterised by concentric condyle–fossa positions
language, which is important for clinical recognition in average range fossa depth/fossa width ratio
in patients, clinical identification of patient subsets shaped fossae. In normals model 2 (19
1%), condyle
and their subsequent prospective testing. This is more position trended to, or was, posterior but also main-
problematic for stepwise logistic regression analysis, tained the average range fossa ratio and shape,
which constructs a single or main model, is not bidi- which presumably is sufficient to maintain a stable
rectional and does not reveal subgroups. disc position and shape despite the slight or definite
Figure 3 graphically illustrates the above three position eccentricity in models 2 and 3. In normals
patterns differentiating subsets of normals (1) that model 3 (19
1%), condyle position trended to ante-
can then be recognised and identified clinically. rior, but still maintained the average range fossa
Normals model 1 was the majority (47
6%) ratio and shape.

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 307

Differentiation of TMJ disc displacement with reduction moderate posterior condyle position combined there-
patients from normals fore with a closer to average posterior joint space sub-
ordinate to the same narrower and or deeper fossa
A classification tree analysis (2) differentiated 3 mod-
shape. 29
8% of DDw/R patients were incorrectly
els of the DDw/R derangement class from normals
classified as normals (false negatives). A logistic
and explained a notable 37 0% of the difference (var-
regression model predicting DDw/R versus normals
iance, R2) orthopedically. The complete model
was by definition only a single model and incorpo-
approached usable sensitivity (70
2%) differentiating
rated greater fossa width, reduced post-glenoid pro-
the disc derangements and excellent specificity
cess height and narrow posterior joint space
(90
5%) differentiating normals.
presumably indicating a posterior condyle position.
Figure 4 illustrates the above three clinically recog-
The model had less prediction value at 67
9% sensi-
nisable TMJ patterns associated with DDw/R defined
tivity and 73
8% specificity, and a lower explained
by the tree (2). The largest group, DDw/R 1 (40
5%),
variance R2 = 20
5% (2). This seems to represent the
was characterised by a much wider and/or shallower
most common classification tree model 1, but also
fossa alone versus normals. Condyle position was
adds in post-glenoid process height which is of biolog-
unspecified or unnecessary at the available sample
ical interest because that was responsive in the
size. Disc displacement with reduction group 2
McNamara (12) experimental orthopaedic histologic
(22
6%) showed an interaction of posterior condyle
adaptation model. Univariate tests were positive for
position and a very narrow posterior joint space sub-
>fossa width, >fossa width/fossa depth ratio reflecting
ordinate to an average to narrower fossa and or a
a wider fossa and added >eminence slope length,
deeper fossa shape. Disc displacement with reduction
probably reflecting the wider fossa (2).
group 3 (7
1%) is a variant of subgroup 2 with only a

Differentiation of disc displacement without reduction from

normals

A classification tree analysis differentiating DDw/oR

TMD patients from normals (2) (Fig. 5) was more
simple at the sample size and constructed 2 models
for this recent onset closed lock sample. It explained a
moderate 28
8% of the variance (R2), and at 66
7%
sensitivity and 85
7% specificity had slightly less pre-
dictive values than the DDw/R group. The largest
DDw/oR Group 1 (58%) (Fig. 5) was characterised by
a posterior to very posterior condyle position alone.
Disc displacement without reduction Group 2 (9%)
had much deeper fossa depths subordinate to more
concentric or trending anterior condyle positions.
Additional fossa characteristics were unspecified in
the locking group at the available sample size. Of the
DDw/oR patients, 33% were incorrectly classified as
normals (false negatives).
The single logistic regression model predicting
DDw/oR versus normals (2) (Fig. 5, model 3) incorpo-
rated deeper fossae, with longer post-glenoid process
height and greater eminence height, greater fossa
width, and greater condyle width, and only margin-
Fig. 4. Clinical illustration of the three osseous anatomic organi-
sations of TM joints with disc displacement with reduction versus ally lower prediction values at 64
1% sensitivity and
normals according to classification tree analysis (models 1–3), 81
0% specificity, and R2 at 24
5%. Univariate analy-
and the single logistic regression-derived model (model 4). sis of isolated variables continued to report the trend

© 2013 Blackwell Publishing Ltd

308 A. PULLINGER

Table 2. Summary of prediction values for multivariate tempo-

DD without romandibular joint anatomic studies
reduction 1
(vs. Normals) Sensitivity Specificity Pseudo
58% (%) (%) R2
Posterior to
very posterior Published guidelines >75 >90 (123)
deeper for TMD
Normals vs. undifferentiated DDw/R + DDw/oR
DD w/o Red 2 Classification tree 67
9 85
7 32
6 (1)
Logistic regression 72
2 76
2 23
3
Normals vs. DDw/R
More concentric or Classification tree 70
2 90
5 37
0 (2)
9%
trending anterior Logistic regression 67
9 73
8 20
5
Normals vs. DDw/oR
deeper
Classification tree 66
7 85
7 28
8 (2)
longer Logistic regression 64
1 81
0 24
5
Logistic Regression
DDw/R vs. DDw/oR
DD w/o Red 3
Classification tree 82
6* 65
4† 31
5 (3)
wider Greater Logistic regression 60
9* 66
7† 9
9
height
DDw/R, disc displacement with reduction; DDw/oR, disc dis-
placement without reduction.
wider *Predicting DDw/oR.
†
Predicting DDw/R.
Fig. 5. Clinical illustration of the two osseous anatomic organi-
sations of TM joints with disc displacement without reduction
versus normals according to classification tree analysis (models 1 Shorter
DD with Reduction 1.
and 2), and the single logistic regression model (model 3). [vs. DD without reduction]
wider

for wider fossae seen in DDw/R (Fig. 4) and brought

in a flatter eminence slope (larger mean radius) and
58%
longer slope length as significant. The fact that these
eminence characteristics were not included in the
final multivariate models, or were unnecessary math-
Shorter
ematically at the sample size, does not necessarily pre- DD w/ Red 2.
wider [vs. DD w/o Red]
clude their biological relevance.
Condyle position entered the classification tree
8%
model at the top tier (2) for DDw/oR (locking group)
with fossa depth measurement subordinate, versus rounder
condyle position being subordinate to fossa width/
depth ratio measurement in the reciprocal clicking
Fig. 6. Clinically illustrates the two osseous anatomic organisa-
(DDw/R) group (Figs 4 and 5), indicating some ortho- tions (models 1 and 2) characterising disc displacement with
paedic differences between the two derangement reduction differences from disc displacement without reduction
classes (Table 2). TM joints.

one model is associated with normal function and also

Conclusions from Figs 2–5
with disc derangement even within the same
There is a relationship between TM joint osseous derangement class, plus the influence of variables can
organisation and function. However, useful prediction be bidirectional. This cannot be appreciated in univar-
is only revealed in multifactor multivariate studies (1, iate analysis, or in stepwise logistic regression models,
2), yet still only explains 20–37% of the variance which although bringing in additional data, results in
(R2). This is substantial statistically, but biologically, at only a single equation or model. Nevertheless, the
best, is only one-third of the explanation. More than regression models were generally compatible with the

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 309

most frequent classification tree model. Logistic noid process combined with a wider superior joint
regression analysis might be more useful on subsets space (Fig. 6). Condyle position and fossa width/depth
identified by classification tree or cluster analysis if were unspecified at the available sample size. A large
numbers permit. superior joint space might not be favourable to main-
taining a stable biconcave disc shape if the superior
and anterior–superior disc thicknesses are more equal,
Disc displacement with reduction joint differences from disc
possibly allowing forward displacement of the disc rel-
displacement without reduction joints
ative to the condyle over time. Group 2 is a less fre-
Differences characterising disc displacement with reduction quent variant of Group 1 (8%) with also a shorter
joints. The classification tree analysis (3) differentiated post-glenoid process but with a wider anterior joint
two anatomic models for the clicking joints differ- space relative to superior joint space and a rounder
ences from the locking joints. The largest DDw/R articular eminence. This suggests a more posterior
Group 1 (58%) was classified by a shorter post-gle- condyle position and unfavourable more open wedge-
shaped joint disc space anteriorly. It is hypothesised
that the curved articular eminence might help resist
longer progression of the forward displacement of the disc or
Disk Displacement
without Reduction 1. its dislocation [see converse in the DDw/oR joints
[vs. DD with reduction] (Fig. 7)]. Linear regression was only run with DDw/
oR as the outcome variable (Fig. 7) and is included
52·2% next.

Differences characterising disc displacement without reduc-

tion joints from joints with disc displacement clicking. The
shorter DD w/o Red 2.
classification tree analysis (3) described two
[vs. DD w/ Red]
organisation differences for DDw/oR joints (Fig. 7)
wider from DDw/R joints. The largest Group 1 shows that a
open 30·4%
longer post-glenoid process alone differentiated most
flatter DDw/oR joints (52
2%). This might imply a longer
preceding period of forward traction stimulation of
the process at the insertion of the articular disc poste-
longer riorly. A similar proliferation of the post-glenoid pro-
DDw/o Red 3 cess was described by McNamara et al. (12) following
(vs. DDw/Red) continuous forward positioning of the condyle in
Logistic regression juvenile monkeys. Condyle position and fossa width/
greater height depth were unspecified at the available sample size.
The second classification tree model for joints with
locking (30
4%) involved an interaction between a
shorter post-glenoid process and a wider more open
longer deeper
anterior joint space plus a flatter articular eminence.
DDw/o Red 4
(vs. DDw/Red)
This suggests a more posterior condyle position and
Univariate unfavourable more wide open wedge-shaped joint
very narrow disc space anteriorly possibly also exaggerated by a
superiorly = flatter articular eminence. The sometimes small ana-
more open anteriorly
tomic organisation differences between DD w/R and
DDw/oR joints might illustrate that small differences
Fig. 7. Clinically illustrates the two osseous anatomic organisa-
can determine very different outcomes. As these were
tions (models 1 and 2) characterising disc displacement without
reduction differences from disc displacement with reduction TM recent onset closed lock cases (1–3), the differences
joints; plus the single logistic regression model (model 3), and a were not due to consequent and subsequent osseous
univariate model 4. remodelling.

© 2013 Blackwell Publishing Ltd

310 A. PULLINGER

The by definition single logistic regression model 3 When DDw/R was compared directly to DDw/oR
(Fig. 7) predicting between DDw/oR and DDw/R had for differences (3), 34
6% of the DD w/R joints were
weaker prediction value at 60
9% sensitivity (predict- classified by the tree as having the anatomic organisa-
ing DDw/oR) and 66
7% specificity (predicting DDw/ tion characteristics predominant to DD w/o reduction
R) with an R2 value of only 9
9%. The greatest differ- joints [false positives].
ences were fossa shape based, incorporating interac- It is therefore asked whether these false-positive
tion of longer post-glenoid process height and greater normals have any increased risk thereby for onset of
eminence height in DDw/oR. Bringing a dynamic disc displacement clicking. It is also asked whether
post-glenoid process into the clinical parameters of the above subset of DDw/R joints has increased risk
condyle–fossa discussion in derangements appears to for symptom progression to locking or perhaps wors-
be new. ening derangement. However, clinical experience and
Univariate comparisons for DDw/oR joints were some data (125) indicate that many fewer than
positive for greater post-glenoid process height, greater 34
6% of joints with reciprocal clicking progress to
fossa depth and a smaller absolute superior joint space, frank closed lock. This underlines the fact that
which could be a proxy for posterior condyle position although the tree model showed differences between
or suggest a wedge-shaped disc anteriorly. DDw/R and DDw/oR (3) that are moderately strong,
Anatomic orthopaedic differences revealed between 68
54% of the variance is still unaccounted for. This
the two main derangement classes (3) (Figs 6 and 7) implies that many other probably non-orthopaedic
indicate these diagnoses do not necessarily belong to factors are operating that need identification and
a single disease continuum (117) differentiated only inclusion in future prospective studies to answer our
by their stage in the process. This may be part of the important risk questions. Conversely, it can be asked
explanation why only a few joints with clicking pro- whether correctly classified normals and DDw/R cases
gress to locking and why only a subset of individuals have any resistance to derangement onset or symp-
with non-problem TMJ noise progress to problem tom progression.
derangements.

The potential functional impact of TM joint organisation on

Summary disc stability

The interaction of condyle–fossa centricity–eccentric-

Risk of TMJ derangement onset and progression
ity with fossa depth and width shape in disc derange-
Clinically, there is obviously a subset that does pro- ments and normals is hypothesised in Fig. 8, in three
gress from benign clicking, to reciprocal clicking, to broad majority conceptual models.
locking, and eventual osteoarthrosis (124). However,
in a 3-year prospective study for progression of TMJ Conceptual model 1 (Fig. 8). Do wider-shallower fossae
reciprocal clicking, only 9% of patients held on an in joints with DDw/R, as in Fig. 4 majority group 1,
extended wait list (125) reported that the clicking dis- lead to increased horizontal vectors of force on the
appeared leaving a limited jaw opening, indicating a disc, disc elongation and laxity, but allow an easy
progression to locking. There was no change in 71% path for the disc to relocate on the condyle with
of the sample, and in fact, the clicking disappeared opening translation, thus avoiding locking/jamming?
and range of motion normalised without deflection in
20% indicating an adaptive return to orthofunction, Conceptual model 2 (Fig. 8). In comparison, normals
although this does not imply a return to a normal disc mostly had approximately concentric condyle posi-
position or shape. tions combined with mid-range fossa width/depth
In the current classification tree models, 9
5% of dimensions. This is an organisation which by defini-
normals had the orthopaedic anatomic characteristics tion is associated with more stable joints, thereby with
that typified the DDw/R disc clicking group [false pos- biconcave disc shapes with more favourable
itives] (2) (Fig. 4) and 14
3% of normals had the biomechanics, along with only average vertical versus
organisation of the pooled DDw/R and DDw/oR horizontal vectors of loading that are anticipated in
patients [false positives] (1) (Fig. 2). mid-range fossa shape and proportions. If not

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 311

1. height and in post-glenoid process height. Thus, it is

Disk Displacement possible that deeper fossae might be found in the
y with Reduction
DDw/oR majority model if expanded to a much larger
x
sample size. Alternatively, if still not participating, it
2. would affirm substantial dichotomy in joints with
Normals
closed lock, possibly with more important non-ortho-
y paedic variables acting. A wider deeper fossa charac-
teristic is identified in the univariate and logistic
xx regression models (Fig. 5).

3.
Disk Displacement
without Reduction
Suggested applications to assessment and
treatment decisions
y
1 There is predictive value in the osseous anatomic
and orthopaedic organisation of the TM joint for
x
normals versus disc derangement subdiagnoses and
Fig. 8. Clinically illustrates the conceptual majority temporo-
between subdiagnoses. This is manifested as signifi-
mandibular joint osseous anatomic models comparing (1) disc cant interactions between condyle–fossa position,
displacement with reduction, (2) normals and (3) disc displace- fossa shape and size variation, and ratios of joint
ment without reduction, discussed in terms of their hypothetical space. However, multifactor analysis and modelling
potential for function–dysfunction. is required for statistical significance and to under-
stand and build clinically recognisable and poten-
concentric, as in Fig. 3 minority normals models 2 tially usable models.
and 3, they still have a normative fossa shape, which 2 The basic biological tenet of a relationship of form
might assist maintaining disc stability. to function is therefore upheld but the prediction
value should not be overstated.
Conceptual model 3 (Fig. 8). Does a deeper fossa depth
in DDw/oR seen in its minority model 2 and the
• Only about one-third of the variance or differ-
ences are explained by these TM joint osseous
logistic regression model 3 (Fig. 5) imply long stand-
anatomic models, implying that two-thirds of
ing differences in the pattern of condyle function? A
the variables are not yet included, and might be
deep fossa may orthopedically affect jaw use, condyle
as or more important, requiring higher priority
and disc loading, and movement patterns with more
in treatment.
traction on the post-glenoid process. More vertical
seating of the condyle might occur in deeper fossae as
• The earliest 1983 American Dental Association
Guidelines (126) are therefore still very perti-
it returns to the closed-packed position with a greater
nent, stating that initial treatment for TMD
vertical vector of force on the posterior part of the
should be generally limited to non-invasive and
disc possibly distorting the posterior band of the
largely reversible modalities. This means that
fibrous part of the disc. This may encourage disc
permanent occlusal and jaw position change
instability and its anterior displacement over time and
should only rarely be used in the first manage-
also make re-engaging the condyle under the displac-
ment of TMD problems and that any invasive or
ing disc more difficult on opening translation, and
non-reversible treatment should only be consid-
certainly once the complete disc becomes dislocated
ered after reassessed for need after control of
forward. The DDw/oR majority model 1 (Fig. 5) did
the acute symptoms.
not specify any fossa data at the available sample size,
3 Although some recognisable anecdotal clinical vari-
instead utilising posterior to very posterior condyle
ables are included or supported, the clinician is
position alone, which differs from the DDw/oR
warned to not use, isolate or extract single variables
minority model 2 that has more concentric positions.
from the multifactor models to apply to diagnosis
However, the overall logistic regression model (Fig. 5)
or treatment. Therefore, posterior condyle position
does incorporate deeper fossa seen both in eminence

© 2013 Blackwell Publishing Ltd

312 A. PULLINGER

alone is not diagnostic and is involved only as a co- 9 Risk avoidance dentistry studies are lacking; hereby
factor in certain derangement subgroups, plus also relatable to minimising the potential physiological
occurs in some normals. Conversely, although adaptation demand in TMJ dysfunction patients
approximate condyle centricity typifies many nor- who require dental treatment. Some dental and
mals, it is also found in many derangements. occlusal treatments have an inherent orthopaedic
4 There is more than one anatomic model that is com- component, which could be helpful or could
patible with normal function and more than one for increase risk in which case should be avoided if a
disc displacement with and without reduction. The different plan is feasible.
contribution of a variable may be bidirectional and
splits samples into subsets. These confound univari-
• The author’s personal advice in cases with TM
joint disc derangements is to (i) avoid any
ate analysis and single logistic regression equations,
strong retrusive forces because this could poten-
as well as clinical diagnostic opinion.
tially exaggerate and aggravate an anteriorly
5 The multifactor nature of jaw dysfunction may be
displacing disc; (ii) similarly avoid equilibration
important biologically because the masticatory sys-
alteration to centric relation occlusion; (iii) if
tem can tolerate extremes in single or few variables
possible favour centric occlusion dentistry to
and still maintain normal function. Hence, the poor
avoid introducing an orthopaedic component;
predictive values when single variables are studied:
(iv) make changes slowly to permit adaptation,
instead requiring a mix of parameters or the wrong
compromised as that might be. (v) In cases
mix to increase risk for dysfunction.
where large changes in guidance or jaw position
• On the other hand, some small anatomic differ- are contemplated, test out on appliances and in
ences could prove critical and might determine provisional restorations. Then, finalise only after
normal function versus TMJ instability and also giving adaptation time and making corrections
progression of disease versus resistance (1–3). that are copied into the permanent restoration.
This is the domain of chaos theory. • Similarly in TMJ active osteoarthritis: (i) try to
6 Temporomandibular joint disc repair surgery seems avoid much condyle position change or do it
problematic in prognosis because this does not slowly; (ii) try to avoid increasing joint loading;
attempt to alter the potential unstable disc environ- (iii) possibly work to a habitual or ‘comfort posi-
ment (fossa shape, or probably condyle position tion’; and (iv) preferably wait until the osteoar-
and associated disc space shape), nor can alter jaw thritis has ‘stabilised’ before making major dental
function habits or other potential soft tissue, or tis- occlusal changes, meanwhile using more provi-
sue type risk factors. sional stabilisation means.
7 Similarly, condyle anterior repositioning (127) • Conversely, most normal joints adapt well to
that tries to orthopedically bring the condyle orthopaedic changes, such as working to centric
position forward therapeutically and in some relation, if appropriate, and vertical dimension
cases permanently forward onto an otherwise changes. However, some adaptation is still
anteriorly displacing disc to stabilise disc function inherent, so big changes should still be tested. It
does not alter the fossa shape environment, or should be remembered that a subset of normal
disc-ligament elongation, or other many tissue joints have some orthopaedic organisation more
factors. typical of the derangement groups, and all dys-
8 Orthodontic treatment. There are still no prospec- functional joints were presumably once normal.
tive scientific data to support bringing a condyle to
a centred position ‘prophylactically’ from a poster-
ior position if the TM joint function is normal. The Future studies
exception may be correction of unilateral functional
1 Substantially increasing sample sizes to bring in
crossbites in children owing to a small risk for
additional osseous orthopaedic variables is only
adaptation at the expense of disc function. The
expected to increase the explained variance slightly,
probable misapplication of the latter to adults is dis-
because the effect in multifactor analysis is relative
cussed earlier in this article.
to the entry variables. However, it might bring in

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 313

additional interesting subsets or better explain • Answers require multifactor models using dis-
existing subsets. ease incidence data instead of cross-sectional
2 TMJ multifactor orthopaedic anatomic models prevalence data.
should be tested in other samples. Some differences • Prospectively following normals identified with
are expected due to population differences and TM joint characteristics of joints with derange-
variables that are more prevalent or not yet studied ment for disease onset is of great interest to
in these models. However, the overall impression study and then apply to informed consent and
and a form-and-function relationship should selection of treatment methodology in ortho-
remain especially in a closed synovial joint system. dontics and dentistry especially those with a
3 The predictive contribution of other potential cofac- potential orthopaedic component. Conversely,
tor subsystems still requires investigation and mod- identification of true negatives that may be
els constructed, for example, disc imaging and more robust functionally is just as important
tissue type, articular lining soft tissue type and sta- clinically. Prospective study will undoubtedly
tus, inflammation and friction (surface tissues and require the addition of several other types of
synovial assays), differences in muscle force and variables. Similarly, it will be important to fol-
vectors (integrates with craniofacial skeletal type), low non-problem disc clicking derangements
pattern of joint loading (integrating with jaw func- with characteristics of joints with locking for
tion characteristics), occlusion extremes, trauma, risk of progression versus resistance.
patterns of parafunction and tension habits affect- • A potential 30-year prospective incidence study
ing loading, oestrogen effects on tissues (noting is difficult to contemplate. However, conducting
how rare complete DDw/oR is seen in males), con- multiple series of 5-year interval studies may be
nective tissue elasticity and laxity, collagen type. possible, then stitch them together statistically
4 Construction of a grand multifactor model has cur- to simulate the long period. A contemporary
rent limitations. short cut may be to try to tissue type DNA gen-

• There seem to be practical limits to the number

ome sequences of groups who have progressed
to derangement versus who have resistance to
of entry variables that the software programs
onset and progression, to apply as markers a
can handle at attainable sample sizes. Reduction
priori, in conjunction with available anatomic
by using interaction variables or indices may be
models, etc.
one approach but risks masking important con-
tributions and subsets versus using the subvari- • Meanwhile, typifying the TM joint organisation
and occlusion in an elderly sample with normal
ables as entry variables.
• Statistically multifactor models are relative to the
TM joints and no history of pain or dysfunction
is the best relevant cross-sectional study that
entry variables used, meaning that strong vari-
has never been carried out (long-term true-neg-
ables (e.g. age, gender, trauma) can mathemati-
ative wellness model).
cally knock out weak variables as unnecessary
mathematically, even though they may in fact be • To make evidence-based decisions, we desper-
ately need prospective outcome studies. It is not
contributing biologically and were identified as
for a lack of numbers considering the number
significant in previous more focused models. For
of practitioners in these fields. Instead, we are
example, it is not a simple matter to add in occlu-
lacking usable database collection.
sion to the TM joint anatomic models because the
6 It is likely that future clinical diagnostic analysis
explained variance will not be simple additive.
will require a computer program to integrate the
Potential new subsystems, for example tissue
multiple variables.
variables, will likely require separate multifacto-
7 Conducting classification tree analysis (or other
rial examination first.
cluster type analysis) together with logistic regres-
5 Prediction of risk for disease onset, derangement
sion is recommended as the norm in all future TMD
progression, prophylactic intervention and contin-
and occlusion studies because each brings a view-
ued normal function are up-front important clinical
point to the table without being contradictory.
questions. These are aetiology questions.

© 2013 Blackwell Publishing Ltd

314 A. PULLINGER

Acknowledgments Growth Series “Moyer’s Symposium”. Univ. Michigan,

Michigan, USA; 1975:209–227.
The author gives full acknowledgement and great 13. Rugh JD, Johnson RW. Vertical dimension discrepancies
thanks to Dr. Donald Seligman DDS as research col- and masticatory pain/dysfunction. In: Solberg WK, Clark
GT, eds. Abnormal jaw mechanics, diagnosis and treat-
league and co-author in most of the UCLA occlusion-
ment. Chicago (IL): Quintessence Inc.; 1984:117–139.
TMD studies. 14. Carlson DS, McNamara JA, eds. Muscle adaptation in the
craniofacial region. Monograph 8, Craniofacial Growth
Series, Center for Human Growth and Development. Ann
References Arbor (MI): University of Michigan; 1978.
1. Pullinger AG, Seligman DA, John MT, Harkins S. Multifac- 15. Lundeen HC, Gibbs CH. Jaw movements and forces dur-
torial modeling of temporomandibular anatomic and ing chewing and swallowing and their clinical signifi-
orthopedic relationships in normal versus undifferentiated cance. In: Lundeen HC, Gibbs CH, eds. Advances in
disk displacement joints. J Prosthet Dent. 2002;87:289– occlusion. Postgraduate Dental handbook. Boston (MA):
297. John Wright PSG Inc.; 1982:2–32.
2. Pullinger AG, Seligman DA, John MT, Harkins S. Multifac- 16. Belser UC, Hannam AG. The influence of altered working-
torial comparison of disk displacement with and without side occlusal guidance on masticatory muscles and related
reduction to normals according to temporomandibular jaw movement. J Prosthet Dent. 1985;53:406–413.
joint hard tissue anatomic relationships. J Prosthet Dent. 17. Le BellY, J€ams€a T, Korri S, Niemi PM, Alanen P. Effect of
2002;87:298–310. artificial occlusal interferences depends on previous expe-
3. Pullinger AG, Seligman DA. Multifactorial analysis of dif- rience of temporomandibular disorders. Acta Odontol
ferences in temporomandibular joint hard tissue anatomic Scand. 2002;60:219–222.
relationships between disk displacement with and without 18. Niemi PM, Le Bell Y, Kylm€al€a M, J€ams€a T, Alanen P. Psy-
reduction in women. J Prosthet Dent. 2001;86:407–419. chological factors and responses to artificial interferences
4. Pullinger A. Establishing better biological models to under- in subjects with and without a history of temporomandib-
stand occlusion. II: occlusal relationships. J Oral Rehabil ular disorders. Acta Odontol Scand. 2006;64:300–305.
(in press). 19. Lerman MD. The muscle engram: the reflex that limits
5. Pullinger AG, Seligman DA, Gornbein JA. A multiple conventional occlusal treatment. Cranio. 2011;29:297–
logistic regression analysis of the risk and relative odds of 303.
temporomandibular disorders as a function of common 20. Celenza FV. The centric position: replacement and charac-
occlusal features. J Dent Res. 1993;72:968–979. ter. J Prosthet Dent. 1973;30:591–598.
6. Pullinger AG, Seligman DA. Quantification and validation 21. Davis SJ, Gray RMJ, Smith PW. Occlusion: occlusal prac-
of predictive values of occlusal variables in temporoman- tice in simple restorative dentistry. Br Dent J.
dibular disorders using a multifactorial analysis. J Prosthet 2001;191:365–381.
Dent. 2000;83:66–75. 22. Popowitch K, Nebbe B, Major PW. Effect of Herbst treat-
7. Seligman DA, Pullinger AG. Analysis of occlusal variables, ment on temporomandibular joint morphology: a system-
dental attrition, and age for distinguishing healthy controls atic literature review. Am J Orthod Dentofac Orthop.
from female patients with intracapsular temporomandibu- 2003;123:388–394.
lar disorders. J Prosthet Dent. 2000;83:76–82. 23. Marsico E, Gatto E, Burrascano M, Matarese G, Cordasco
8. Fantoni F, Chiappe G, Landi N, Romagnoli M, Bosco M. A G. Effectiveness of orthodontic treatment with functional
stepwise multiple regression model to assess the odds ratio appliances on mandibular growth in the short term. Am J
between myofascial pain and 13 occlusal features in 238 Orthod Dentofacial Orthop. 2011;139:24–36.
Italian women. Quintessence Int. 2010;41:e54–e61. 24. Myers D, Barenie J, Bell R, Williamson E. Condylar posi-
9. Chiappe G, Fantoni F, Landi N, Biondi K, Bosco M. Clini- tion in children with functional posterior crossbites before
cal value of 12 occlusal features for the prediction of disc and after crossbite correction. Pediatr Dent. 1980;2:190–
displacement with reduction (RDC/TMD Axis I group IIa). 194.
J Oral Rehabil. 2009;36:322–329. 25. Troelstrup B, Moller E. Electromyography of the temporal-
10. Landi N, Manfredini D, Tognini F, Romagnoli M, Bosco is and masseter muscles in children with unilateral cross-
M. Quantification of the relative risk of multiple occlusal bite. Scand J Dent Res. 1970;78:425–430.
variables for muscle disorders of the stomatognathic sys- 26. Ingervall D, Thilander D. Activity of the temporal and
tem. J Prosthet Dent. 2004;92:190–195. masseter muscles in children with lateral forced bite.
11. Kaifu Y, Kasai K, Townsend G, Richards L. Tooth wear Angle Orthod. 1975;45:249–258.
and the design of the human dentition: a perspective from 27. Andrade AS, Gameiro GH, Derossi M, Gaviao MB. Poster-
evolutionary medicine. Yearb Phys Anthropol. 2003;46: ior crossbite and functional changes. A systematic review.
47–61. Angle Orthod. 2009;79:380–386.
12. McNamara JA. Determinants of mandibular form and 28. Hesse KL, Artun J, Joondeph DR, Kennedy DB. Changes
growth. In: McNamara JA, ed. Monograph 4. Craniofacial in condyle position and occlusion associated with maxil-

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 315

lary expansion for correction of functional unilateral pos- 44. Kawakami M, Yamamoto K, Inoue M, Kawakami T, Fu-
terior crossbite. Am J Orthod Dentofacial Orthop. jimoto M, Kirita T. Morphological differences in the tem-
1997;111:410–418. poromandibular joints in asymmetrical prognathism
29. Kilic N, Kiki A, Oktay H. Condylar asymmetry in unilat- patients. Orthod Craniofac Res. 2006;9:71–76.
eral posterior crossbite patients. Am J Orthod Dentofacial 45. Ueki K, Nakagawa K, Takatsuka S, Shimada M, Marukawa
Orthop. 2008;133:382–387. K, Takazakura D et al. Temporomandibular joint morphol-
30. Thurston MH. Craniofacial characteristics associated ogy and disc position in skeletal class III patients. J Cra-
with unilateral crossbite in the permanent dentition, MS niomaxillofac Surg. 2000;28:362–368.
Oral Biology Thesis. University of California Los Angeles; 46. Nickel JC, McLachlan KR, Smith DM. A theoretical model
1986. of loading and eminence development of the postnatal
31. Pullinger A, Thurston M, Turley P. Condyle adaptation to human temporomandibular joint. J Dent Res. 1988;67:
unilateral posterior cross bite. J Dent Res. 1985;64:269. 903–910.
32. O’Byrn BL, Sadowsky C, Schneider B, BeGole EA. An 47. Katsavrias EG. Changes in articular eminence inclination
evaluation of mandibular asymmetry in adults with uni- during the craniofacial growth period. Angle Orthod.
lateral posterior crossbite. Am J Orthod Dentofacial Ort- 2002;72:258–264.
hop. 1995;107:394–400. 48. Lawther LL. A roentgenographic study of the temporo-
33. Langberg BJ, Arai K, Miner RM. Transverse skeletal and mandibular joint using a special head positioner. Angle
dental asymmetry in adults with unilateral posterior cross- Orthod. 1956;26:22–33.
bite. Am J Orthod Dentofac Orthop. 2005;127:15–16. 49. Csad
o K, M
arton K, Kivovics P. Anatomical changes in the
34. Pellizoni SE, Salioni MA, Juliano Y, Guimaraes AS, Alonso structure of the temporomandibular joint caused by com-
LG. Temporomandibular joint disc position and configura- plete edentulousness. Gerodontology. 2012;29:111–116.
tion in children with functional unilateral posterior cross 50. Raustia AM, Pirttiniemi P, Salonen MA, Pyhtinen J. Effect
bite: a magnetic resonance imaging evaluation. Am J Or- of edentulousness on mandibular size and condyle-fossa
thod Dentofacial Orthop. 2006;129:785–793. position. J Oral Rehabil. 1998;25:174–179.
35. Thilander B, Bjerklin K. Posterior crossbite and temporo- 51. Pullinger AG, Baldioceda F, Bibb CA. Relationship of TMJ
mandibular disorders (TMDs): need for orthodontic treat- articular soft tissue to underlying bone in young adult
ment? Eur J Orthod. 2012;34:667–673. condyles. J Dent Res. 1990;69:1512–1518.
36. Tecco S, Festa F. Prevalence of signs and symptoms of 52. The Academy of prosthodontics. The glossary of prosth-
temporomandibular disorders in children and adolescents odontic terms. J Prosthet Dent. 1999;81:39–110.
with and without crossbites. World J Orthod. 2010;11: 53. McKee JR. Comparing condylar positions achieved
37–42. through bimanual manipulation to condylar positions
37. Arat FE, Arat ZM, Tompson B, Tanju S. Muscular and achieved through masticatory muscle contraction against
condylar response to rapid maxillary expansion. Part 3: an anterior deprogrammer: a pilot study. J Prosthet Dent.
magnetic resonance assessment of condyle-disc relation- 2005;94:389–393.
ship. Am J Orthod Dentofacial Orthop. 2008;133:830–836. 54. O’Ryan F, Epker BN. Temporomandibular joint function
38. Farella M, Michelotti A, Iodice G, Milani S, Martina R. and morphology: observations on the spectra of normalcy.
Unilateral posterior crossbite is not associated with TMJ Oral Surg Oral Med Oral Pathol. 1984;58:272–279.
clicking in young adolescents. J Dent Res. 2007;86:137– 55. Hannam AG. Current computational modelling trends in
141. craniomandibular biomechanics and their clinical implica-
39. Pullinger AG, Seligman DS, Solberg WK. Temporomandib- tions. J Oral Rehabil. 2011;38:217–234.
ular disorders. Part II: occlusal factors associated with tem- 56. Nickel JC, McLachlan KR. In vitro measurement of the
poromandibular joint tenderness and dysfunction. J stress-distribution properties of the pig temporomandibu-
Prosthet Dent. 1988;59:363–367. lar joint disc. Arch Oral Biol. 1994;39:439–448.
40. Solberg WK, Bibb CA, Nordstrom BB, Hansson TL. Maloc- 57. Nickel JC, Iwasaki LR, Beatty MW, Moss MA, Marx DB.
clusion associated with temporomandibular joint changes Static and dynamic loading effects on temporomandibular
in young adults at autopsy. Am J Orthod. 1986;89:326– joint disc tractional forces. J Dent Res. 2006;85:809–813.
330. 58. Iwasaki LR, Crosby MJ, Gonzalez Y, McCall WD, Marx
41. Akahame Y, Deguchi T, Hunt N. Morphology of the tem- DB, Ohrbach R et al. Temporomandibular joint loads in
poromandibular joint in skeletal Class III symmetrical and subjects with and without disc displacement. Orthop Rev
asymmetrical cases: a study by cephalometric analysis. J (Pavia). 2009;1:90–93.
Orthod. 2001;28:119–128. 59. Pullinger AG, Seligman DA. Overbite and overjet charac-
42. Pirttiniemi P, Raustia A, Kantomaa T, Pyhtinen J. Rela- teristics of refined diagnostic groups of temporomandibu-
tionships of bicondylar position to occlusal asymmetry. lar disorder patients. Am J Orthod Dentofac Orthop.
Eur J Orthod. 1991;13:441–445. 1991;100:401–415.
43. Wang MQ, He JJ, Chen CS, Widmalm SE. A preliminary 60. Weinberg LA. Optimum temporomandibular joint condyle
anatomical study on the association of condylar and position in clinical practice. Int J Periodontics Restorative
occlusal asymmetry. Cranio. 2011;29:111–116. Dent. 1985;5:10–27.

© 2013 Blackwell Publishing Ltd

316 A. PULLINGER

61. Gelb H. The optimum temporomandibular joint condyle condylar head and displacement of the temporomandibu-
position in clinical practice. Int J Periodontics Restorative lar joint disk. Dentomaxillofac Radiol. 2001;30:162–165.
Dent. 1985;5:34–61. 78. Ikeda K, Kawamura A. Disc displacement and changes in
62. Dawson PE. Optimum TMJ condyle position in clinical condylar position. Dentomaxillofacial Radiol. 2012;1–8,
practice. Int J Periodontics Restorative Dent. 1985;5:10–31. doi:10.1259/dmfr/84227642.
63. McNeill C. The optimum temporomandibular joint con- 79. Berry DC, Watkinson AC. Mandibular dysfunction and
dyle position in clinical practice. Int J Periodontics Restor- incisor relationship: a theoretical explanation for the click-
ative Dent. 1985;5:52–76. ing joint. Br Dent J. 1978;144:74–77.
64. Thomas CL, ed. Taber’s cyclopedic medical dictionary, 21st 80. Owen AH. Orthodontic/orthopedic treatment of cranio-
edn. Philadelphia (PA): FA Davis Co., 2012. mandibular pain dysfunction. Part 2: posterior condylar
65. Dawson PE. A classification system for occlusions that displacement. J Craniomandibular Pract. 1984;2:333–349.
relates maximal intercuspation to the position and condi- 81. Mongini F. Anatomic and clinical evaluation of the rela-
tion of the temporomandibular joints. J Prosthet Dent. tionship between the temporomandibular joint and occlu-
1996;75:60–66. sion. J Prosthet Dent. 1977;38:539–551.
66. Gonzalez B. The not-so-controversial issue of condylar 82. Sonnesen L, Svensson P. Temporomandibular disorders
position. Int J Orthod Milwaukee. 2007;18:17–26. and psychological status in adult patients with a deep bite.
67. Zhao YP. Evaluation of the normal adults’ condylar posi- Eur J Orthod. 2008;30:621–629.
tion in the fossa. Zhonghua Kou Qiang Yi Xue Za Zhi. 83. Carlsson GE, Egermark I, Magnusson T. Predictors of signs
1993;28:70–72, 127. and symptoms of temporomandibular disorders: a 20-year
68. Pullinger AG, Hollender L, Solberg WK, Petersson A. A follow-up study from childhood to adulthood. Acta Odon-
tomographic study of mandibular condyle position in an tol Scand. 2002;60:180–185.
asymptomatic population. J Prosthet Dent. 1985;53:706– 84. Zhou D, Hu M, Liang D, Zhao G, Liu A. Relationship
713. between fossa-condylar position, meniscus position, and
69. Vitral RW, da Silva Campos MJ, Rodrigues AF, Fraga MR. morphologic change in patients with Class II and III mal-
Temporomandibular joint and normal occlusion: is there occlusion. Chin J Dent Res. 1999;2:45–49.
anything singular about it? A computed tomographic 85. Gianelly AA, Petras JC, Boffa J. Condylar position and
examination. Am J Orthod Dentofacial Orthop. Class II deep-bite, no-overjet malocclusions. Am J Orthod
2011;140:18–24. Dentofac Orthop. 1989;96:428–432.
70. Vukovojac S, Krmpotic I, Kovacevic D. Radiographic anal- 86. Cohlmia JT, Ghosh J, Sinha PK, Nanda RS, Currier GF.
ysis of condyle position in patients with TMJ dysfunctions. Tomographic assessment of temporomandibular joints in
Acta Stomatol Croat. 1990;24:97–109. patients with malocclusion. Angle Orthod. 1996;66:
71. Incesu L, Taskaya-Yilmaz N, Ogutcen-Toller M, Uzun E. 27–35.
Relationship of condyle position to disk position and mor- 87. Pullinger AG, Solberg WK, Hollander L, Petersson A. Rela-
phology. Eur J Radiol. 2004;51:269–273. tionship of mandibular condylar position to dental occlu-
72. Pullinger AG, Solberg WK, Hollander L, Guichet D. Tomo- sion factors in an asymptomatic population. Am J Orthod
graphic analysis of mandibular condyle position in diag- Dentofacial Orthop. 1987;91:200–206.
nostic subgroups of temporomandibular disorders. J 88. Artun J, Hollender LG, Trulove EL. Relationship between
Prosthet Dent. 1986;55:723–729. orthodontic treatment, condylar position, and internal
73. Pereira LJ, Gaviao MB, Bonjardim LR, Castelo PM. Ultra- derangement in the temporomandibular joint. Am J Or-
sound and tomographic evaluation of temporomandibular thod Dentofacial Orthop. 1992;101:48–53.
joints in adolescents with and without signs and symp- 89. Gianelly AA, Cozzani M, Boffa J. Condylar position and
toms of temporomandibular disorders: a pilot study. maxillary first premolar extraction. Am J Orthod Dentofa-
Dentomaxillofac Radiol. 2007;36:402–408. cial Orthop. 1991;99:473–476.
74. Ren YF, Isberg A, Westesson PL. Condyle position in the 90. Gianelly AA, Anderson CK, Boffa J. Longitudinal evalua-
temporomandibular joint. Comparison between asymp- tion of condylar position in extraction and nonextraction
tomatic volunteers with normal disk position and patients treatment. Am J Orthod Dentofacial Orthop.
with disk displacement. Oral Surg Oral Med Oral Pathol 1991;100:416–420.
Oral Radiol Endod. 1995;80:101–107. 91. Egermark I, Carlsson GE, Magnusson T. A prospective long-
75. Abdel-Fattah RA. Optimum temporomandibular joint term study of signs and symptoms of temporomandibular
(TMJ) condylar position. Todays FDA. 1989;1:1C–3C. disorders in patients who received orthodontic treatment in
76. Ozawa S, Boering G, Kawata T, Tanimoto K, Tanne K. childhood. Angle Orthod. 2005;75:645–650.
Reconsideration of the TMJ condylar position during 92. McNamara JA, Seligman DA, Okeson JP. Orthodontic
internal derangement: comparison between condylar posi- treatment, and temporomandibular disorders: a review. J
tion on tomogram and degree of disk displacement on Orofac Pain. 1995;9:73–90.
MRI. Cranio. 1999;17:93–100. 93. Liu ZJ, Yamagata K, Kuroe K, Suenaga S, Noikura T, Ito
77. Kurita GH, Ohtsuka A, Kobayashi H, Kurashina K. A G. Morphological and positional assessments of TMJ com-
study of the relationship between the position of the ponents and lateral pterygoid muscle in relation to symp-

© 2013 Blackwell Publishing Ltd

 TM JOINT ANATOMIC RELATIONSHIP PREDICTING FUNCTION 317

toms and occlusion of patients with temporomandibular of the lateral pterygoid muscle and variation in angulation
disorders. J Oral Rehabil. 2000;27:860–874. of the medial pterygoid muscle: mandibular mechanics
94. Manfredini D. Etiopathogenesis of disk displacement of implications. J Prosthet Dent. 1988;60:616–621.
the temporomandibular joint: a review of the mecha- 108. Taskaya-Yilmaz N, Ceylan G, Incesu L, Muglaili M. A pos-
nisms. Indian J Dent Res. 2009;20:212–221. sible etiology of the internal derangement of the temporo-
95. Stegenga B. Osteoarthritis of the temporomandibular joint mandibular joint based on the MRI observations of the
organ and its relationship to disc displacement. J Orofac lateral pterygoid muscle. Surg Radiol Anat. 2005;27:19–
Pain. 2001;15:193–205. 24.
96. Hasegawa H, Saitoh I, Nakakura-Ohshima K, Shigeta K, 109. Bittar GT, Bibb CA, Pullinger AG. Histological characteris-
Yoshihara T, Suenaga S et al. Condylar shape in relation tics of the lateral pterygoid muscle insertion to the tempo-
to anterior disc displacement in juvenile females. Cranio. romandibular joint. J Orofac Pain. 1994;8:243–249.
2011;29:100–110. 110. Gibbs CH, Mahan PE, Wilkinson TM, Mauderli A. EMG
97. Bryndahl F, Warfvinge G, Eriksson L, Isberg A. Cartilage activity of the superior belly of the lateral pterygoid mus-
changes link retrognathic growth to TMJ disc displace- cle in relation to other jaw muscles. J Prosthet Dent.
ment in a rabbit model. Int J Oral Maxillofac Surg. 1984;51:691–701.
2011;40:621–627. 111. Wilkinson TM. The relationship between the disk and the
98. Gidarakou IK, Tallents RH, Kyrkanides S, Stein S, Moss lateral pterygoid muscle in the human temporomandibu-
ME. Comparison of skeletal and dental morphology in lar joint. J Prosthet Dent. 1988;60:715–724.
asymptomatic volunteers and symptomatic patients with 112. Davies SJ. Malocclusion-a term in need of dropping or
bilateral disk displacement without reduction. Angle Or- redefinition? Br Dent J. 2007;202:519–520.
thod. 2003;73:684–690. 113. Selaimen CM, Jeronymo JC, Brilhante DP, Lima EM,
99. Gidarakou IK, Tallents RH, Kyrkanides S, Stein S, Moss Grossi PK, Grossi ML. Occlusal risk factors for temporo-
ME. Comparison of skeletal and dental morphology in mandibular disorders. Angle Orthod. 2007;77:471–477.
asymptomatic volunteers and symptomatic patients with 114. Seren E, Akan H, Toller MO, Akyar S. An evaluation of
normal temporomandibular joints. Angle Orthod. the condylar position of the temporomandibular joint by
2003;73:116–120. computerized tomography in Class III malocclusions: a
100. Gidarakou IK, Tallents RH, Kyrkanides S, Stein S, Moss preliminary study. Am J Orthod Dentofacial Orthop.
M. Comparison of skeletal and dental morphology in 1994;105:483–488.
asymptomatic volunteers and symptomatic patients with 115. Katsavrias EG, Halazonetis DJ. Condyle and fossa shape in
bilateral degenerative joint disease. Angle Orthod. Class II and Class III skeletal patterns: a morphometric
2003;73:71–78. tomographic study. Am J Orthod Dentofacial Orthop.
101. Badel T, Panduric J, Marotti M, Kern J, Krolo I. Metric 2005;128:337–346.
analysis of temporomandibular joint in asymptomatic per- 116. Liedberg J, Rohlin M, Westesson PL. Observer perfor-
sons by magnetic resonance imaging. Acta Med Croatica. mance in assessment of condylar position in temporoman-
2008;62:455–460. dibular joint radiograms. Acta Odontol Scand. 1985;43:
102. Haiter-Neto F, Hollender L, Barclay P, Maravilla KR. Disk 53–58.
position and the bilaminar zone of the temporomandibu- 117. Pullinger A, Hollender L. Variation in condyle-fossa rela-
lar joint in asymptomatic young individuals by magnetic tionships according to different methods of evaluation in
resonance imaging. Oral Surg Oral Med Oral Pathol Oral tomograms. Oral Surg. 1986;62:719–727.
Radiol Endod. 2002;94:372–378. 118. Epstein JB, Caldwell J, Black G. The utility of panoramic
103. Ribeiro RF, Tallents RH, Katzberg RW, Murphy WC, Moss imaging of the temporomandibular joint in patients with
ME, Magalhaes AC et al. The prevalence of disc displace- temporomandibular disorders. Oral Surg Oral Med Oral
ment in symptomatic and asymptomatic volunteers age 6 Pathol Oral Radiol Endod. 2001;92:236–239.
to 25 years. J Orofac Pain. 1997;11:37–47. 119. Pullinger A, Hollender L. Assessment of mandibular con-
104. Rammelsberg P, Pospiech PR, Jager L, Pho Duc JM, Bohm dyle position: comparison of transcranial radiographs and
AO, Gernet W. Variability of disc position in asymptom- linear tomograms. Oral Surg. 1985;60:329–334.
atic volunteers and patients with internal derangements 120. Stamm T, Hohoff A, Van Meegen A, Meyer U. On the
of the TMJ. Oral Surg Oral Med Oral Pathol Oral Radiol three-dimensional physiological position of the temporo-
Endod. 1997;83:393–399. mandibular joint. J Orofac Orthop. 2004;65:280–289.
105. Westesson P-L, Erikson L. Condylar and disc movements 121. Ikeda K, Kawamura A, Ikeda R. Assessment of optimal
in dissected TMJ autopsy specimens, video. Lund, Swe- condylar position in the coronal and axial planes with
den: University of Lund, 1985. limited cone-beam computed tomography. J Prosthodont.
106. Juniper RP. The shape of the condyle and position of the 2011;20:432–438.
meniscus in temporomandibular joint dysfunction. Br J 122. Endo M, Terajima M, Goto TK, Tokumori K, Takahashi I.
Oral Maxillofac Surg. 1994;32:71–76. Three-dimensional analysis of the temporomandibular
107. Christiansen EL, Roberts D, Kopp S, Thompson JR. CT joint and fossa-condyle relationship. Orthodontics (Chic.).
assisted evaluation of variation in length and angulation 2011;12:210–221.

© 2013 Blackwell Publishing Ltd

318 A. PULLINGER

123. Widmer CG, Lund JP, Feine JS. Evaluation of diagnostic romandibular disorders. J Am Dent Assoc. 1983;106:
tests for TMD. J Calif Dent Assoc. 1990;18:53–60. 75–77.
124. Rasmussen OC. Clinical findings during the course of tem- 127. Eberhard D, Bantleon HP, Steger W. The efficacy of ante-
poromandibular arthropathy. Scand J Dent Res. rior repositioning splint therapy studied by magnetic reso-
1981;89:283–288. nance imaging. Eur J Orthod. 2002;24:343–352.
125. Lundh H, Westesson PL, Kopp S. A three year follow-up
of patients with reciprocal clicking. Oral Surg Oral Med Correspondence: Andrew Pullinger, UCLA School of Dentistry, CHS
Oral Pathol. 1987;63:530–533. 43-045, 10833 Le Conte Ave, Los Angeles, CA 90024-1668, USA.
126. Griffiths RH. Report on the president’s conference on E-mail: apullinger@dentistry.ucla.edu
the examination, diagnosis, and management of tempo-

© 2013 Blackwell Publishing Ltd