See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/11181484

The curve of Spee and craniofacial morphology:

A multiple regression analysis
Article in European Journal Of Oral Sciences September 2002
DOI: 10.1034/j.1600-0722.2002.21255.x Source: PubMed

CITATIONS

READS

30

478

4 authors, including:
Mauro Farella

Ambra Michelotti

University of Otago

University of Naples Federico II

141 PUBLICATIONS 1,622 CITATIONS

120 PUBLICATIONS 1,644 CITATIONS

SEE PROFILE

SEE PROFILE

Roberto Martina
University of Naples Federico II
98 PUBLICATIONS 918 CITATIONS
SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

Available from: Mauro Farella

Retrieved on: 16 September 2016

Copyright
Eur J Oral Sci 2002

Eur J Oral Sci 2002; 110: 277281

Printed in UK. All rights reserved

European Journal of
Oral Sciences
ISSN 0909-8836

The curve of Spee and craniofacial

morphology: a multiple regression
analysis

Mauro Farella1, Ambra Michelotti1,

Theo M. G. J. van Eijden2,
Roberto Martina1
1

Department of Orthodontics, University of

Naples Federico II, Naples, Italy, 2Department
of Functional Anatomy, Academic Center for
Dentistry Amsterdam (ACTA), Amsterdam, The
Netherlands

Farella M, Michelotti A, van Eijden TMGJ, Martina R. The curve of Spee and
craniofacial morphology: a multiple regression analysis. Eur J Oral Sci 2002; 110:
277281.
Eur J Oral Sci, 2002
The aim of this study was to investigate the relationship between the curve of Spee and
skeletal facial morphology. Dental casts and lateral cephalograms were obtained from
59 orthodontic patients. The amount of concavity of the curve of Spee was calculated
by a second-order quadratic interpolation of buccal cusp tips obtained from lateral
digital photographs of the teeth. The cephalometric analysis aimed to evaluate the
sagittal and vertical craniofacial dimensions as well as the position of the mandibular
condyle with respect to the occlusal plane. These variables, included in a multiple
regression model, could explain 34% of the total variance of the curve of Spee. The
amount of the curvature was signicantly related to (a) the horizontal position of the
condyle with respect to the dentition, (b) the sagittal position of the mandible with
respect to the anterior cranial base, and (c) the ratio between the posterior and anterior
facial height. No signicant relationship was found between the curve of Spee and any
of the other cephalometric variables. The curve of Spee was not inuenced by age and
gender of the subjects investigated.

From a lateral view, the occlusal plane in adult human

dentition is generally concave with the concavity facing
upward. The curve of Spee (1, 2) is dened in the glossary of prosthodontics terms as the anatomic curve established by the occlusal alignment of the teeth, as
projected onto the median plane, beginning with the cusp
tip of the mandibular canine and following the buccal
cusp tips of the premolar and molar teeth (3).
The functional signicance of the curvature has not
been completely understood. The morphological arrangement of the teeth in the sagittal plane has been
related to the slope of the articular eminence, incisor
overbite, molar cusp height, and the amount of posterior
contact; a matched interaction between these features
and the curve of Spee ensures balanced occlusal function
(4). More recently, it has also been suggested that the
curve of Spee has a biomechanical function during food
processing by increasing the crushshear ratio between
the posterior teeth and the eciency of occlusal forces
during mastication (5, 6).
Leveling of the curve of Spee represents a routine
procedure in orthodontic practice (7), and the more
pronounced the curve is, the greater the additional space
required to atten the dentition (8). The assessment of
the depth of the curve of Spee is therefore a critical point
for the orthodontic diagnostic and treatment protocol
(9).
Surprisingly, the inuence of craniofacial morphology
on the curve of Spee has been systematically investigated
in very few studies and with conicting ndings (1012).

Mauro Farella, Department of Orthodontics,

University of Naples Via S. Pansini 5, I)80131
Naples, Italy
Telefax: +39-81-7462192
E-mail: Farella@unina.it
Key words: curve of Spee; craniofacial
morphology; occlusal plane.
Accepted for publication April 2002

In those studies, a simple correlation analysis between

the radius or the depth of the curve of Spee and skeletal
facial types was conducted. This approach, however,
ignores the simultaneous contribution of multiple factors
to the individual variation of the curve. Therefore, the
aim of the present study was to determine the relationship between the curve of Spee and several features of
craniofacial morphology by means of a multiple regression analysis. The inuences of gender and age on the
amount of the curvature were also investigated.

Material and methods

Sample and clinical records
Complete pretreatment records of 59 patients were selected
from the record store of the Orthodontic Department at the
University of Naples Federico II, Italy. The sample
investigated (16 males, 43 females; mean age SD
20.3 5.6 yr; range 1542 yr) consisted of a group with
dierent degrees of vertical and sagittal skeletal discrepancies. All the subjects were white Caucasian adults from
Campania, Italy.
Inclusion criteria were a complete permanent dentition
(with the exception of third molars) and an age of more than
15 yr for females and more than 18 yr for males. The
patients were selected from an initial sample of 836
according to the following additional exclusion criteria:
previous orthodontic treatment; unilateral or bilateral
posterior cross-bite; periodontal disease; posterior dental
crowding in the lower arch (from canine to second molar);

278

Farella et al.

morphological tooth anomaly; dental restoration or crown;

mandibular asymmetry; self-reported symptoms of
temporomandibular disorders.
Dental casts were reproduced in white stone (Orthodontic
Stone; Whip Mix, Louisville, KY, USA) from alginate
impressions (Alginoplast; Bayer, South Bend, IN, USA).
Lateral standardized cephalograms were taken by a single
technician using the same X-ray device. The cephalograms
were made with the mandible in the intercuspal position.
The focus to midsagittal plane distance was 152 cm, and the
lm to midsagittal plane distance was 10 cm. The values
measured were not corrected for linear enlargement
(approximately 7% in the median plane).
Measurement of the curve of Spee
The left hand-side of the mandibular dental cast was photographed by means of a digital camera (CoolPix 900;
Nikon, Tokyo, Japan) xed in a photo copystand. The cast
was oriented in a way that the lens of the camera was parallel to the buccal surfaces of the posterior teeth. The line
between the distal cups of the second molar and the cusp of
the canine was oriented parallel to the horizontal axis of the
camera display. The camera-to-teeth distance was 10 cm,
and the magnication was 1 : 1 for all the photographs.
On each photograph, the cusp tips of the molars, premolars and canine were marked and were entered in a
Cartesian axis reference (Fig. 1). The x-axis was superimposed on the line connecting the distal cusp of the second
molar to the cusp of the canine, the origin was at the midpoint of this line, and the y-axis was perpendicular to x-axis
through the origin. Digitized coordinates of dental cusps
were interpolated with a best-t second-order polynomial
function y a + bx + cx2. The curve of Spee (Sp) was
calculated as c 104. This coecient indicates how much
the curve is concave: the larger it is, the more concave the
curve. When Sp is equal to 0, the curve is a straight line. The
concavity of the curve is upward when the sign of Sp is
positive and downward when the sign is negative.
Digitization of cusp tips, calculation of Cartesian
coordinates and interpolation of the curve were performed
by a single examiner by means of specic software [Adobe
Photoshop 5.0, Microsoft Excel 97, Jones TechDig, 1.1
(Ronald Jones, IL, USA)]. The procedure for the measurement of the curve of Spee was similar to that described
by Ferrario et al. (13).

Cephalometric measurements
The cephalometric analysis aimed to evaluate the sagittal
and vertical craniofacial dimensions. Vertical and
horizontal cephalometric measurements are shown in
Fig. 2a,b. The percentage ratio between the posterior facial
height (S-Go) and the anterior total facial height (N-Me),
and the percentage ratio between the anterior upper facial
height (N-ANS) and the anterior lower facial height (ANSMe) were also calculated. Since it has been reported that the
amount of the curve of Spee is related to the position of the
dentition with respect to the mandibular condyle (5, 6, 14),
this position was also included in the measurements
(Fig. 2c). A single examiner performed all cephalometric
measurements.
Error of the method
The errors of the method were calculated from 25 randomly
selected subjects. A second examiner retraced the radiograms as well as the digital pictures for the calculation of Sp.
The method error (ME) for all these measurements was
assessed by means of Dahlbergs formula

s
P 2
d
ME
2n
where d is the dierence between the two measurements and
n is the number of recordings (15). Systematic dierences
between replicate measurements were tested with paired
Students t-test setting the a error at 0.1 (16). Mean errors
were, in general, low. The errors for the distances varied
between 0.2 mm and 1.1 mm, and for the angles between
0.1 and 1.0. The mean error of Sp was 0.8 and ranged
from 0 to 2.1. There was no systematic error for either
cephalometric measurements or the coecient Sp (Students
t-test P > 0.1).
Statistics
Data were analysed with conventional descriptive statistics.
Mean, standard deviation (SD), standard error of mean
(SEM), range, and coecients of variation (CV
SD/mean 100) were calculated for the coecient Sp and
for all the cephalometric variables. The normal distribution
tting of data collected was tested by means of KolmogorovSmirnov test. The hypothesis that data collected were
normally distributed could not be rejected for any variable,
therefore subsequent analysis were performed by means of
parametric statistical tests. The goodness of t of Sp was
calculated by correlation coecients. A multiple linear
regression analyses with a stepwise elimination was
performed. The coecient Sp was considered as the
response variable. Explanatory variables included in the
model were gender, age, and the cephalometric measures
previously described. All the analyses were carried out by
means of commercial statistical software (SPSS 8.0; SPSS,
Chicago, IL, USA). Statistical signicance was accepted at
P < 0.05.

Results
Fig. 1. Cartesian axis references used for the measurement of
the curve of Spee.

Descriptive statistics for the cephalometric measurements are given in Table 1. The correlation coecients of

Curve of Spee and craniofacial morphology

279

signicant relationship was found between the curve of

Spee and all the other cephalometric measurements. The
curve of Spee was not signicantly inuenced by the age
and the gender of the subjects investigated.

Discussion
The mathematical and statistical approach used in the
present study allows the description of the intrinsic shape
of the mandibular curve of Spee. Consistent with previous observations (13), in most cases the second-order
interpolation curve was judged to t the cusp tips very
well. Therefore, it appeared appropriate for the description of the curve.
As revealed by the coecient of variation of the curve
of Spee (Sp 42%), the amount of the curvature was
highly variable among the individuals. Nevertheless, in
only one case was the curve found to be almost completely at (Sp 2), and in no case was the coecient 0
(straight line) or negative (reversed curve).
In the subjects investigated, the amount of curvature
was not inuenced by gender. The lack of sexual dimorphism is in agreement with previous ndings obtained from either two-dimensional (10, 13) or threedimensional (17) evaluation of the occlusal curvature. In
the present study, no evidence was found about the potential relation between the curve of Spee and age. The
eect of age on the curve of Spee has been investigated in
growing subjects up to young adulthood (1820), however, to our knowledge, there is no data available regarding variation of the curve of Spee with age in adults.
With increasing age, it can be expected that a homogeneous dental wear may occur as a consequence of the
masticatory function. It is interesting to note that the
curve of Spee is usually seen in most of the primates and
mammals regardless of the amount of tooth wear (5, 6).
Tooth wear does not appear to aect the forward tilt of
the posterior teeth in the sagittal plane and even though

Sp ranged from 0.78 to 0.99 (mean SD 0.91 0.06).

After the stepwise elimination, the total variance of the
curve of Spee that was explained by the regression model
was 34%. The adjusted R2 was 30.1%.
The amount of the curvature was signicantly related
to the horizontal position of the dentition with respect to
condyle (O-M), to the sagittal position of the mandible
with respect to the anterior cranial base (SNB), and to
the percentage ratio between anterior and posterior facial height (S-Go : N-Me percentage) (Table 2). No

Fig. 2. (a) Vertical measurements: NSL/ANS-PNS, angle

between sella-nasion line and palatal line; NSL/ML, angle
between sella-nasion line and mandibular line; ANS-PNS/ML,
angle between palatal line and mandibular line; Ar-Go-Me,
angle between the ramal line and mandibular line; S-Go, distance between sella and gonion; N-Me, distance between nasion
and menton; N-ANS, distance between nasion and anterior
nasal spine, ANS-Me: distance between anterior nasal spine
and menton. (b) Horizontal measurements: S-N, distance
between sella and nasion; SNA, angle between sella-nasion line
and nasion-A line; SNB, angle between sella-nasion line and
nasion-B line; ANB, angle between nasion-A line and nasion-B
line; Cd-Gn, distance between condylion and gnathion; Cd-A,
distance between condylion and point A. (c) Position of lower
dentition with respect to condyle: condyle point (J), the center
of the articular surface of the condyle; the molar point (M), the
most posterior point on the occlusal line of the lower dentition;
the incisor point (I), the most anterior point on the occlusal line
of the lower dentition; the O point (O), intersection between the
occlusal line (IM) and the perpendicular to IM through J.
The horizontal position of M was dened with the distance
OM, the vertical position with the distance OJ. These
measurements were assessed according to Osborn (5).

280

Farella et al.
Table 1
Summary statistics for the curve of Spee and cephalometric variables (n 59)*

Curve of Spee (Sp)

Cephalometric variables
NSL/ANS-PNS ()
NSL/ML ()
ANS-PNS/ML ()
Ar-Go-Me ()
S-Go (mm)
N-Me (mm)
S-Go/N-Me (%)
N-ANS/ANS-Me (%)
S-N (mm)
SNA ()
SNB ()
ANB ()
Cd-Gn (mm)
Cd-A (mm)
O-M (mm)
O-J (mm)

Mean

SD

SEM

10.6

4.5

0.6

8.9
31.4
23.1
125.5
83.0
122.4
67.9
80.1
71.4
81.5
77.8
3.6
120.8
92.3
42.1
39.0

3.7
7.7
7.0
6.9
7.7
9.4
6.0
8.6
4.3
4.4
4.3
2.9
8.2
5.8
5.3
5.6

0.5
1.0
0.9
0.9
1.0
1.2
0.8
1.1
0.6
0.6
0.6
0.4
1.1
0.8
0.7
0.7

Minimum

Maximum

CV (%)

2.0

23.0

42.0

0.0
15.0
10.0
113.0
71.0
105.0
54.1
62.7
60.5
70.0
66.5
)4.0
103.0
78.0
31.0
28.5

15.0
50.0
42.0
144.0
104.0
143.0
80.2
100.0
83.0
94.0
88.5
11.0
143.0
104.0
53.0
50.0

41.1
24.6
30.5
5.5
9.3
7.6
8.8
10.7
6.0
5.4
5.5
80.5
6.8
6.2
12.5
14.3

*NSL/ANS-PNS, angle between sella-nasion line and palatal line; NSL/ML, angle between sella-nasion line and mandibular line;
ANS-PNS/ML, angle between palatal line and mandibular line; Ar-Go-Me, angle between the ramal line and mandibular line; S-Go,
distance between sella and gonion; N-Me, distance between nasion and menton; N-ANS, distance between nasion and anterior nasal
spine, ANS-Me, distance between anterior nasal spine and menton; S-N, distance between sella and nasion; SNA, angle between
sella-nasion line and nasion-A line; SNB, angle between sella-nasion line and nasion-B line; ANB, angle between nasion-A line and
nasion-B line; Cd-Gn, distance between condylion and gnathion; Cd-A, distance between condylion and point A; O-M, distance
between point O and point M; O-J, distance between point O and point J (see Fig. 2).
Table 2
Stepwise multiple regression analysis for the curve of Spee*
Dependent
variable
Sp

Statistically signicant
independent variables

SE B

P-value

R2
change

Constant
O-M
S-Go/N-Me %
SNB

24.9
)0.25
0.34
)0.34

9.4
0.1
0.1
0.1

0.01
0.01
0.001
0.02

0.20
0.07
0.07

*Sp is the curve of Spee. B is the unstandardized regression coecient. SE B is the standard error of B.

the cusps become more and more at, the curve of Spee
is maintained throughout the years.
It has been suggested that the position of the mandibular condyle in relation to the occlusal plane has a
considerable eect on the patterns of movements of the
lower teeth when the jaws are closed. These patterns
aect the way in which food is processed during mastication (crushshear ratio). In a theoretical model, Baragar & Osborn (6) have shown that an increased
forward tilting of the posterior lower teeth (i.e. curve of
Spee more marked) is mechanically advantageous when
the J point is raised up or moved anteriorly with respect
to the dentition. The sagittal arrangement of the teeth
with the curve of Spee may represent a strategy to align
the relatively large posterior forces that occur during
mastication along the long axes of the molar and premolar teeth. Page (14) summarized a geometrical description of this concept in the tangent law.
The ndings of the present study support the hypothesis that the position of the mandibular condyle in

relation to the occlusal plane inuences the shape of the

curve of Spee. More specically, the more the lower
dentition is located forward (with respect to condyle), the
less marked is the curve. The sagittal position of the
lower dentition explained about 20% of the variance of
the curve of Spee.
The curve of Spee was also inuenced by the position
of the mandible with respect to the anterior cranial base
(i.e. SNB angle), regardless of the reciprocal position of
the lower and upper jaw in the sagittal plane (ANB). The
more the mandible is positioned anteriorly, the less
marked is the curve. The variance explained by the SNB
angle was, however, rather limited, amounting to about
7%. The lack of correlation between ANB angle and the
curve of Spee contrasts with previous reports from
Orthlieb (12), who found that the radius of the curve of
Spee was shorter in third class than in second class
malocclusions.
When considering the vertical measurements assessed
in the present study, the curve of Spee was signicantly

Curve of Spee and craniofacial morphology

inuenced by the ratio between the posterior and anterior

facial height. This variable is indicative for the vertical
craniofacial morphology. The multiple regression analysis suggests that the curve of Spee is more marked in
short-face subjects and less marked in long-face subjects.
Nevertheless, it should be stressed that the ratio between
posterior and anterior facial height explained only 7% of
the total variance of the curve. A potential explanation
for the slight dierences in the curve of Spee between
short-face and long-face subjects may be ascribed to their
dierent occlusal forces, with the short-face subject exhibiting higher forces (21, 22).
Overall, the ndings of the present study suggest that
the stomatognathic system adjusts to dentoskeletal
variation by varying the shape of the curve of Spee, but
to a minor extent. Indeed, only one-third of the variance
of the curve could be explained by the cephalometric
variables included in the regression model. Most of the
variation of the curvature remains unexplained by craniofacial morphology and could probably be explained
better by other factors. It is noteworthy that a positive
correlation between the curve of Spee and the inclination
of the masseter muscle has been reported (23). It would
be interesting to evaluate the simultaneous contribution
of several muscular lines of action to variation of the
curve.
The management of the sagittal organization of the
teeth represents a critical point for the stability of
complete dentures, and perhaps it is also implicated in
the failures of implant-supported restorations. More
insight into the morphological and functional signicance of the curve of Spee might also improve the
stability of orthodontic treatment, particularly in the
correction of deep-bite malocclusions; further studies
should therefore be encouraged.
It may be concluded that the curve of Spee is highly
variable between individuals and is inuenced only to a
minor extent by craniofacial morphology. Variation of
the curve was mostly explained by the horizontal position of the mandibular condyle in relation to the occlusal
plane. The amount of the curvature was weakly inuenced by the vertical craniofacial dimensions and by the
position of the mandible with respect to the anterior
cranial base. The curve of Spee is independent of the
gender of the individual, and no evidence was found for a
relation between age and the amount of curvature in
adult subjects.

281

References
1. Spee FG. Die Verschiebungsbahn des Unterkiefers am Schadel.
Arch Fur Anat und Physiol Leipzieg. Verlag Veit und Comp,
1890, 285293.
2. Spee FG. The gliding path of the mandible along the skull. J
Am Dent Assoc 1980; 100: 670675.
3. Academy of Prosthodontics. Glossary of prosthodontic
terms. J Prosthet Dent 1994; 71: 50112.
4. Ramfjord SP, Ash MM. Occlusion. Philadelphia:
WB Saunders 1983.
5. Osborn JW. Relationship between the mandibular condyle and
the occlusal plane during hominid evolution: some of its effects
on jaw mechanics. Am J Phys Anthropol 1987; 73: 193207.
6. Baragar FA, Osborn JW. Efficiency as a predictor of human
jaw design in the sagittal plane. J Biomech 1987; 20: 447457.
7. Tweed CH. Clinical orthodontics. St Louis: CV Mosby, 1966.
8. Braun S, Hnat WP, Johnson BE. The curve of Spee revisited.
Am J Orthod Dentofacial Orthop 1996; 110: 206210.
9. Germane N, Staggers JA, Rubenstein L, Revere JT. Arch
length considerations due to the curve of Spee: a mathematical
model. Am J Orthod Dentofacial Orthop 1992; 102: 251255.
10. Orthlieb JD, Slaviceck R. Geometrische interpretation der
Spee kurve. Z Stomatologie 1985; 82: 118.
11. Ghezzi F, Drago E, De Thomatis P, Zallio F. Profondita`
della curva di Spee in rapporto a FMA, FH, ANB e classe
dentale. Mondo Ortodontico 1991; 1: 7376.
12. Orthlieb JD. The curve of Spee: understanding the sagittal
organization of mandibular teeth. Cranio 1997; 15: 333340.
13. Ferrario VF, Sforza C, Miani A, Colombo A, Tartaglia G.
Mathematical definition of the curve of Spee in permanent
healthy dentitions in man. Arch Oral Biol 1992; 37: 691694.
14. Page HL. The occlusal curve. Dent Digest 1952; 58: 1921.
15. Dahlberg G. Statistical methods for medical and biological
students. New York: Interscience Publications 1940.
16. Houston WJ. The analysis of errors in orthodontic measurements. Am J Orthod 1983; 83: 382390.
17. Ferrario VF, Sforza C, Miani A. Statistical evaluation of
Monsons sphere in healthy permanent dentitions in man. Arch
Oral Biol 1997; 42: 365369.
18. Ferrario VF, Sforza C, Poggio CE, Serrao G, Colombo A.
Three-dimensional dental arch curvature in human adolescents
and adults. Am J Orthod 1999; 115: 401405.
19. Enlow DH. Normal variations in facial form and the anatomic
basis for malocclusion. In: Enlow, DH, ed. Facial growth, 3rd
edn. Philadelphia: WB Saunders, 1990, 193221.
20. Osborn JW, Francis LJ. The position of the dentition in the
mandible and its possible relation to orthodontic abnormalities.
Am J Orthod 1989; 96: 327332.
21. Proffit WR, Fields HW, Nixon WL. Occlusal forces in normal- and long-face adults. J Dent Res 1981; 62: 566571.
22. Ringqvist M. Isometric bite force and its relation to dimensions of the facial skeleton. Acta Odontol Scand 1973; 31: 3542.
23. Osborn JW. Orientation of the masseter muscle and the curve
of Spee in relation to crushing forces on the molar teeth of
primates. Am J Phys Anthropol 1993; 92: 99106.