d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/dema

Influence of cavity dimension and restoration methods on

the cusp deflection of premolars in composite restoration

Mi-Ra Lee, Byeong-Hoon Cho, Ho-Hyun Son, Chung-Moon Um, In-Bog Lee ∗
Department of Conservative Dentistry and Dental Research Institute, College of Dentistry, Seoul National University,
28-2 Yeongeon-Dong, Jongro-Gu, Seoul 110-749, South Korea

a r t i c l e i n f o a b s t r a c t

Article history: Objectives. The aim of this study was to measure the cusp deflection by polymerization
Received 19 April 2005 shrinkage during composite restoration for mesio-occluso-distal (MOD) cavities in premo-
Received in revised form lars, and to examine the influence of cavity dimension, C-factor and restoration method on
12 December 2005 the cusp deflection.
Accepted 10 January 2006 Methods. Thirty extracted maxillary premolars of similar size were prepared with four differ-
ent sizes of MOD cavity, and divided into six groups. The width and depth of the pulpal wall
of the cavity were as follows: group 1: 1.5 × 1 mm, group 2: 1.5 × 2 mm, group 3: 3 × 1 mm,
Keywords: and groups 4–6: 3 × 2 mm. Groups 1–4 were restored using a bulk filling with a composite.
Cusp deflection Group 5 was restored incrementally, and group 6 was restored with an indirect composite
Polymerization shrinkage inlay. Cusp deflections were measured using LVDT transducers. The cusp deflections were
Composites compared between groups using ANOVA and Scheffe tests, and a correlation analysis was
C-factor done.
Bulk filling Results. The cusp deflections of groups 1–4 were 12.1 (2.2), 17.2 (1.9), 16.2 (0.8) and 26.4
Incremental filling (4.2) ␮m, respectively. There was a strong positive correlation between the length cubed
Indirect composite inlay divided by the thickness cubed of the remaining cusp (L3 /T3 ) and cusp deflection. The C-
factor was related to the % flexure (100 × cusp deflection/cavity width). The cusp deflections
of groups 5 and 6 were 17.4 (2.0) and 17.9 (1.4) ␮m, respectively, which were much lower
values than those of group 4.
Significance. : The cusp deflection increased with increasing cavity dimension and C-factor.
Use of an incremental filling technique or an indirect composite inlay restoration could
reduce the cuspal strain.
© 2006 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

1. Introduction
Recently, posterior composite restoration has become more
common because of patients’ increased demand for esthetic
restoration, improvement of adhesive dentistry due to the
dentin bonding system, and concern about amalgam toxicity.
However, the major drawback of composite restoration is
the high polymerization shrinkage of the composite material.
It has been reported that linear shrinkage ranges from 0.2 to
2% [1–3] and that volumetric shrinkage ranges from 0.9 to 5.7%
[3–7], using in vitro measurements.
Polymerization shrinkage can lead to detachment of the
restoration from the tooth surface, or may induce enamel
microcracks. As a consequence, secondary caries and post-
operative hypersensitivity due to bacterial infiltration via
microleakage may occur [8,9]. In addition, when the bond

∗
Corresponding author. Tel.: +82 2 2072 3953; fax: +82 2 2072 3859.
E-mail address: inboglee@snu.ac.kr (I.B. Lee).
0109-5641/$ – see front matter © 2006 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.dental.2006.01.025
 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295
strength between the adhesive and the tooth is strong enough,
the tooth structure may experience shrinkage stress, resulting
in a cusp deflection [10–17].
Cusp deflection is the result of interactions between
the polymerization shrinkage stress of the composite and
the compliance of the cavity wall, and is a common
biomechanical phenomenon observed in teeth restored with
composites.
In order to measure cusp deflection, many meth-
ods have been developed, involving photography [10],
microscopy [11,12], strain gauge [13,14], interferometery
[15] and linear variable differential transformer (LVDT)
[13,16,17]. Cusp deflection during composite restoration
has been reported to be about 10–45 ␮m, varying accord-
ing to the measurement method, tooth type and cavity
size.
There are two important categories of biomechanical fac-
tors that influence cusp deflection. The first category is
composed of geometric and material factors, such as cav-
ity width, cavity depth [10,13,16], the thickness of remain-
ing tooth material [15,18], the polymerization shrinkage of
the composite [15], flow [19] and the compliance of cured
composite and tooth [11,15]. The second category is com-
prised of clinical factors, such as use of liner [11,20], filling
technique (bulk cure versus incremental cure) [10,14,21–23],
restoration methods (direct versus indirect) [24] and use of a
light curing method, which influences the polymerization rate
[25,26].
Hood [18] reported that the remaining cusp after cavity
preparation acts as cantilever beams under occlusal load. The
prepared cavity floor serves as a fulcrum for cusp bending; the
cantilever length is increased with cavity depth. According to
mechanical principles, the cusp deflection is proportional to
the cantilever length cubed, and to the inverse of the thickness
of the cantilever cusp cubed.
Feilzer et al. [27,28] and Davidson and Feilzer [29] reported
that the extent of polymerization shrinkage stress can be
influenced by the cavity configuration (C-factor, bonded sur-
face/unbonded free surface). As the C-factor increases, the
compensation for polymerization shrinkage by the flow of
composite decreases, and thus, the polymerization stress at
the bonded surface increases.
Many researchers have suggested an incremental filling
technique for composite restoration to reduce polymeriza-
tion shrinkage stress and cusp deflection [10,14,29]. How-
ever, controversy remains over whether incremental filling
can reduce cusp flexure as compared to bulk filling. Segura
and Donly [10] and McCullock and Smith [14] reported that
the cusp deflection of incrementally filled teeth was signifi-
cantly lower than that of bulk filled teeth. However, Versluis
et al. [21] and Abbas et al. [22] reported that an incremental
filling technique generated more shrinkage stress. In addi-
tion, Rees et al. [23] reported that there was no significant
difference in cusp flexure between the bulk and incremental
placement.
The use of indirect composite inlay restoration has recently
begun to increase. Indirect restoration was expected to
improve the physical properties of restoration and to result
in less shrinkage stress, because bulk polymerization occurs
extraorally and the space for resin cement is very thin [24].
289
However, polymerization shrinkage stress is affected by the
cavity configuration as well as the extent of polymerization
shrinkage of the composite itself [28]. When the compos-
ite inlay is bonded, the high C-factor with few free surfaces
cannot compensate for the polymerization shrinkage stress.
As a consequence, the remaining stress causes cusp deflec-
tion and microcracks in the tooth [30]. Rees and Jacobsen [24]
reported a cusp tip movement of 0.2–7.4 ␮m using compos-
ite inlay restoration. However, most studies mentioned above
have only reported the measurement values of cusp deflec-
tion, with few biomechanical analyses of the factors affecting
the cusp deflection.
The purpose of this study was to measure the cusp deflec-
tions of premolars restored with composite by bulk filling for
four different dimensions of MOD cavities, and to compare
the cusp flexure resulting from the bulk filling with that of the
incremental filling and indirect composite inlay. This biome-
chanical analysis of the results provides a guideline for suc-
cessful composite restoration in clinic.
2. Materials and methods
2.1. Instrumentation for a measurement device
The device used for measuring cuspal deflection consists of
two XYZ tables (Micro Motion Technology, Bucheon, Korea)
with three attached micrometers (Mitutoyo, Kawasaki, Japan)
and two LVDT probes (AX-1, Solartron Metrology, West Sus-
sex, UK) (Fig. 1). The LVDT probes are capable of detect-
ing linear changes in a range of ±1 mm with a resolution
less than 0.1 ␮m. The calibration of the probe was adjusted
to 10 V/mm (10 mV/␮m) of the output voltage using the
micrometer.
Cuspal deflection was detected by the LVDT, and the mea-
sured value (as a function of time) was stored on a computer
through a data acquisition board, PCI-6024 (National instru-
ment, Mopac Expwy, Austin, TX, USA), using a data acquisition
and analysis software, Labview (National instrument, Mopac
Expwy, Austin, TX, USA).
2.2. Preparation of teeth
Thirty maxillary premolars extracted for orthodontic treat-
ment were stored in chloramine-T solution. The teeth were
divided into six groups of similar average bucco-lingual width.
The dimensions of the teeth for each group are shown in
Table 1. Each tooth was buried 3 mm below the cemento-
enamel junction in an acrylic mold with dimensions of 15 mm
Fig. 1 – Configuration of the instrument for measuring cusp
deflection.
290 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295

Table 1 – The dimension of the teeth and prepared cavity sizes for each group in this study
Group Teeth dimension Cavity size Restoration
method
B-L (mm) M-D (mm) Width (mm) Depth (mm)

1 9.79 (0.39) 7.48 (0.62) 1.5 1.0 Bulk filling

2 9.74 (0.54) 7.74 (0.55) 1.5 2.0 Bulk filling
3 9.72 (0.51) 7.73 (0.36) 3.0 1.0 Bulk filling
4 9.76 (0.45) 7.61 (0.49) 3.0 2.0 Bulk filling
5 9.96 (0.45) 7.41 (0.56) 3.0 2.0 Incremental
6 9.87 (0.50) 7.68 (0.44) 3.0 2.0 Indirect inlay

B-L: bucco-ligual dimension, M-D: mesio-distal dimension.

internal diameter, 25 mm external diameter, and 20 mm
height, using yellow stone.
The teeth set in the acrylic mold were fixed with a
vice and four different sizes of MOD cavities were prepared
using a high speed turbine and a diamond bur. The width
and depth of the pulpal wall of the MOD cavities were
as follows: group 1: 1.5 × 1 mm, group 2: 1.5 × 2 mm, group
3: 3 × 1 mm, and groups 4–6: 3 × 2 mm (Table 1). The high
speed turbine was set on a xyz-table with micrometers and
was able to prepare each cavity precisely. The mesio-distal
proximal box was extended 0.5 mm bucco-lingually, and the
widths of the axial and gingival walls of the box were 1 mm
(Fig. 2a). The reference point for cavity depth was the central
groove.
Fig. 2 – Diagram of prepared cavity. (a) The dimensions of
the occlusal cavity and proximal box (W: width of cavity, D:
depth of cavity, L: length of remaining cusp). (b) Simplified
cavity shape for calculation of the C-factor.
Groups 1–4 were restored using the bulk filling technique,
and groups 5 and 6 were restored using incremental filling and
indirect composite inlay, respectively.
2.3. Measurement of cuspal deflection
Prepared cavity was etched with a 37% phosphoric acid
etchant (Scotchbond, 3 M Co., St. Paul, MN, USA) for 15 s, rinsed
with water and blot dried to leave moist (total etching and wet
bonding). Primer and adhesive (Scotchbond Multipurpose, 3M
Co., St. Paul, MN, USA) were applied according to the manufac-
turer’s instructions and light cured with the Elipar Freelight 2
(3M Co., St. Paul, MN, USA) for 10 s. Thereafter, the specimens
were fixed using a metal vice, and the LVDT probes were posi-
tioned at the buccal and lingual cusp tips (Fig. 1).
Groups 1–4 were bulk filled with Filtek Z-250 composite (3M
Co., St. Paul, MN, USA) and light cured initially at occlusal sur-
face for 40 s, then the light guide was tilted 45◦ to mesial side
and light cured for 40 s, and finally the light guide tilted 45◦ dis-
tally and light cured for 40 s. The total curing time was 120 s.
The additional cure at the mesial and distal direction was to
confirm the good polymerization.
Group 5 was incrementally filled with four layers of the
same composite. The proximal box was first filled to the height
of the pulpal floor and the rest of the cavity was horizontally
filled in three successive increments. Each incremental layer
was light cured for 20 s; first layer was filled and light cured for
20 s, then second layer was filled and cured for 20 s, and these
procedures were repeated for third and fourth layer, and addi-
tionally cured for 40 s. The total cure time was 120 s, the same
as groups 1–4.
For group 6, indirect composite inlays were made using
a Tescera ATL system (Bisco Inc., Itasca, IL, USA). The inlays
were sandblasted and treated with a silane primer (Rely X, 3 M
Co., St. Paul, MN, USA) and adhesive according to the manu-
facturer’s instructions. The adhesive resin applied onto the
prepared cavity and resin inlay was not cured to prevent the
inlay from improper seating. After seating the inlay into the
cavity with an adhesive resin cement (Rely X ARC, 3M Co., St.
Paul, MN, USA), excess resin cement was removed and light
cured with the same protocol as groups 1–4. The total curing
time was also 120 s.
The buccal and lingual cusp movements were recorded for
10,000 s and the total amount of each cuspal movement was
calculated. The room temperature during measurement was
25 ± 0.5 ◦ C. The storage of the teeth, preparation of the cavity
 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295 291

% Flexure (%)
and measurement were undertaken with a special care to pre-

Standard deviations are in parenthesis. % Flexure = 100 × (cusp deflection/cavity width)%. Same letters within the cusp deflection column are not significantly different (ANOVA and Scheffe test,
vent from dehydration of the tooth.

0.81 (0.15)
1.15 (0.12)
0.54 (0.03)
0.88 (0.14)

0.58 (0.07)

0.6 (0.05)
10,000 s
Table 2 – The remaining cusp dimensions (length; L, thickness; T), the L3 /T3 , C-factor, cusp deflection measured, and % flexure according to restoration methods and
2.4. Analysis of the measurements

In order to analyze the relationship between the geometric

factors of the cavity preparation and the cuspal deflection, the

16.2 (0.8)b,c
26.4 (4.2)d
12.1 (2.2)a
17.2 (1.9)c

17.4 (2.0)c

17.9 (1.4)c
Cusp deflection (␮m)
10,000 s
thickness (T) and length (L) of the remaining cusps after cavity
preparation were measured, and the length cubed divided by
the thickness cubed (L3 /T3 ) was calculated (Table 2). The thick-
ness (T) of one remaining cusp was obtained by subtracting the
cavity width from the bucco-lingual width of each tooth and
divided by half. The length (L) of the cusp was determined by

11.3 (2.2)
16.4 (1.7)
15.8 (0.8)
25.4 (4.1)

15.1 (1.4)

17.8 (1.4)
1500 s
averaging the measured distances from the pulpal wall of the
cavity to the buccal and lingual cusp tips (Fig. 2a).
The shape of the cavity was simplified into a solid model
composed of rectangles, as in Fig. 2b, and the C-factor was cal-

2, 3, 4th increment; 1.19

culated from the ratio of the bonded surface to the unbonded

1st increment; 1.25

surface for each tooth using the mesio-distal width, the bucco-

1.67 (0.03)
1.96 (0.04)
1.31 (0.01)
1.42 (0.02)
C-factor
lingual width, and the depth of the prepared cavity. In the case
of the incrementally filled group, it was supposed that the
first increment of composite for the proximal box was filled
at the level of the pulpal floor, and the remaining portion was
equally layered with three successive horizontal increments
(Table 2).
The ratio of the cuspal deflection to the bucco-lingual width

0.46 (0.15)
1.02 (0.29)
0.91 (0.19)
1.74 (0.32)
of the cavity was defined as the percentage of flexure (% flex- L3 /T3
ure). The % flexure was calculated from 100 times the mea-
sured cuspal deflection divided by the bucco-lingual width of
the prepared cavity. The % flexure was to examine the effect
of the C-factor on the cuspal deflection. The relationships
between the cuspal deflection, L3 /T3 , % flexure and C-factor
Remaining cusp dimension (mm)

Length (L)

were analyzed.
3.17 (0.50)
4.10 (0.19)
3.18 (0.35)
4.04 (0.04)

4.13 (0.20)

4.12 (0.17)
The data were analyzed by one-way ANOVA and Scheffe
tests at a significance level of 95% for multiple comparisons
using SPSS version 11.0 software. Pearson’s correlation test
was also used to analyze factors influencing the cuspal deflec-
tion.
Thickness (T)

4.15 (0.19)
4.12 (0.27)
3.36 (0.25)
3.38 (0.22)

3.48 (0.23)

3.44 (0.25)

3. Results

The representative curves of cusp deflection as a function of

time for bulk filling of groups 1–4 are shown in Fig. 3a. The
cusp deflections mainly occurred within the initial 1500 s and
reached a plateau after 2000 s.
1.5 × 1 (group 1)
1.5 × 2 (group 2)
Cavity size

3 × 1 (group 3)
3 × 2 (group 4)

3 × 2 (group 5)

3 × 2 (group 6)

Mean cusp deflection for group 1 was 12.1 ␮m at 10,000 s

W × D (mm)

and those of groups 2, 3, and 4 were 17.2, 16.2 and 26.4 ␮m,
respectively (Table 2, Fig. 4). The amount of cusp deflection
increased as the L3 /T3 increased (p < 0.05), and there was no
significant difference between groups 2 and 3 (p > 0.05).
A comparison of the restoration methods used for groups
Indirect composite inlay

4–6 showed that the cusp deflection of incremental filling

Restoration method

(17.4 ␮m) and indirect composite inlay (17.9 ␮m) were signifi-
Incremental-cure

cantly lower than that of bulk filling (26.4 ␮m). There was no
cavity sizes

significant difference between groups 5 and 6 (p > 0.05) (Table 2,

Fig. 4).
Bulk-Cure

p > 0.05).

When an incremental filling method was used, the cusp

deflection increased in a stepwise manner for each incremen-
tal layer. The cuspal deflection was slightly decreased during
292 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295
Fig. 3 – Representative curves of cusp deflection as a
function of time. (a) Bulk cure (groups 1–4). (b) Bulk cure vs.
incremental cure and indirect composite inlay (groups 4–6).
each light irradiation, but increased again after light exposure
(Fig. 3b).
In the indirect composite inlay, the cusp deflection slowly
increased without light exposure after cementation, and
increased rapidly with the initiation of light curing, before
reaching an initial plateau. After completion of light curing,
cusp deflection again increased rapidly and reached a final
plateau (Fig. 3b).
Fig. 4 – Mean cusp deflections for each tested group.
Groups not statistically different (ANOVA and Scheffe test,
p > 0.05) are denoted with the same letter.
Table 3 – The results of correlation analysis between the
L3 /T3 , C-factor and cusp deflection or % flexure for
groups 1–4
L3 /T3 (R) C-factor (R)
Cusp deflection 0.827* −0.239
% Flexure 0.218 0.808*
∗
Correlation is significant at the 0.01 level.
Fig. 5 – Presentation of correlations between the L3 /T3 ,
C-factor and cusp deflection or % flexure for group 1–4. (a) A
positive correlation was present between the L3 /T3 and the
cusp deflection. (b) A positive correlation was present
between the C-factor and the % flexure.
Statistical analysis showed that there was a strong positive
correlation between the C-factor and the % flexure (R = 0.808),
and between the L3 /T3 and the cusp deflection (R = 0.827,
Table 3, Fig. 5a and b).
4. Discussion
In this study, the change in cusp deflection according to cavity
dimension was investigated by altering the width and depth
of prepared cavities. Unterbrink and Liebenberg [20] reported
that shrinkage stress increases with increasing C-factor, and
that the size of the restored cavity acts as an important factor
in bulk filling. Hood [18] proposed that the cusps remaining
after cavity preparation behave as cantilever beams under
occlusal loads. The ideal deformation of a cantilever beam,
 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295
according to its length, width, and thickness, is given by
[18,31]:
load × length3
deformation =
4 × elasticmodulus × width × thickness3
Hood [18] also described that although both cusp shape and
cavity design are much more complex than a simple cantilever
beam, the concept could be applied to predict the cusp deflec-
tion according to the width and depth of prepared cavity. In the
equation, the force inducing cusp deflection is polymerization
shrinkage during composite curing. The force depends on the
polymerization shrinkage, the total amount of composite in
a cavity, and the elastic modulus of the cured composite. In
addition, as the cavity width and depth increase, the thickness
of the remaining cusp decreases, and the length of cusp wall
increases. The polymerization shrinkage force is also affected
by the flow of composite and the C-factor. It is very difficult
to establish a formula estimating the cusp deflection from so
many variables; therefore, it is practical to measure deflection
experimentally.
There was a significant difference in cusp deflection
according to cavity width and depth seen in this study. The
polymerization shrinkage and elastic modulus of cured com-
posite are identical for each group because the same compos-
ite was used. If the elastic modulus and mesio-distal dimen-
sion of teeth are identical, the cusp deflection may be propor-
tional to the ratio of the length cubed to the thickness cubed
of the remaining cusp (L3 /T3 ). In this study, the L3 /T3 was cal-
culated for each tooth. The averages of L3 /T3 for groups 1–4
are shown in Table 2. There was a strong positive correlation
between the L3 /T3 and the cusp deflection (R = 0.827). There-
fore, cusp deflection may be predicted from the length and
thickness of the remaining cusp.
The difference between groups 2 and 3 cannot be sim-
ply compared using only the cavity dimension and measured
cusp deflection. For comparison, the percent ratio of the cusp
deflection to the cavity width (% flexure), should be consid-
ered in addition to cavity dimension. The mean cusp deflec-
tions for group 2 (1.5 × 2 mm) and group 3 (3 × 1 mm) were
17.2 (±1.9) ␮m and 16.2 (±0.8) ␮m, respectively. There was no
statistical difference between these two groups. However, the
% flexure of group 2 was 1.15%; nearly double that of group
3, 0.54%. That is, the narrow and deep cavity of group 2
showed a much higher % flexure. This could be explained by
an increased C-factor, which indicates diminished composite
flow to compensate for shrinkage, thus increasing the shrink-
age stress at the bonded surfaces, resulting in greater cusp
deflection.
In this study, the C-factors for all cavities in groups 1–5
were calculated. The average C-factor in the groups using bulk
filling was least in group 3, followed by groups 4, 1 and 2,
in ascending order. In the incrementally filled group 5, the
C-factor of the first layer was 1.25, and each of the three suc-
cessive layers had a C-factor of 1.19 (Table 2).
The C-factor showed a significant correlation with the %
flexure (R = 0.808), rather than cusp deflection. In particular,
there was no significant difference in cusp deflection between
groups 2 and 3. However, the C-factors of groups 2 and 3 were
1.96 and 1.31, respectively. This indicates that the narrow and
deep cavity of group 2 has less free surface to compensate for
293
polymerization shrinkage stress with flow. As a consequence,
group 2 showed a high % flexure. Therefore, the amount of
cusp deflection can be explained not only by cavity dimension,
but also by considering the C-factor and % flexure simultane-
ously.
The cusp deflection of the incrementally filled group (group
5) was significantly less than that of bulk filled group (p < 0.01).
This result was consistent with that of other experiments.
Segura and Donly [10] reported cusp deflections of 9.8–12.4 ␮m
and Abbas et al. [22] reported cusp deflections of 17.8–34.0 ␮m
with similar incremental filling methods. Rees et al. [23]
reported a mean cusp deflection of 11.3 ␮m. It is difficult to
compare these results directly, however, because these experi-
ments differ according to cavity sizes, measurement methods,
and type of composites used.
In the incrementally filled group (group 5), the first incre-
ment of composite was filled to the height of the proxi-
mal box and the remaining cavity was filled by three equal
increments. When restored using this method, the calculated
C-factor of the first increment was 1.25, the C-factors for
the 2nd–4th increments were 1.19. These were lower than
the measurement of 1.42 for the C-factor of the bulk filled
group (group 4). The cusp deflection of the incrementally
filled group was decreased by 34.1% compared to that of the
bulk filled group. Therefore, for cavities of the same size, the
incremental filling method lowers the C-factor, which has
the advantage of reducing polymerization shrinkage stress.
This method is also more favorable than bulk cure because
of good composite adaptation to cavity wall, a decrease
in microleakage and an increased degree of conversion
[23].
The cusp deflection of the indirect composite inlay group
(group 6) was also significantly lower than that of bulk filled
group (p < 0.01). The cusp deflection seen in this study, 17.9 ␮m,
was higher than the result from Rees and Jacobsen [24], which
used finite element analysis.
The advantages of indirect composite inlay are as follows
[32]. First, the degree of conversion of cured composite inlay
is increased by pressure, heat and oxygen-free polymerization
in laboratory. Second, the contour of proximal surfaces can be
restored exactly. Third, reduced cement space may reduce the
polymerization shrinkage.
The stress generated during bonding composite inlay may
be related to the cavity shape, compliance of teeth, shrinkage
of resin cement and the thickness of the resin cement. Alster
et al. [33] reported that as the resin cement space increases,
the shrinkage stress is decreased, and that this stress is influ-
enced by the compliance of teeth. However, Rees and Jacobsen
[24] reported that as the thickness of resin cement increases,
the tensile stress of the resin inlay increases to 0.4–17.0 MPa.
Feilzer et al. [27] reported that the linear polymerization
shrinkage of the thin bonded resin layer approximates vol-
umetric polymerization shrinkage as the C-factor increases.
Therefore, when bonding the inlays or crown, increased stress
can cause postoperative pain, cohesive failure, and marginal
gap formation.
We have measured the cement spaces using a SEM in an
additional experiment. The set specimens of group 6 (resin
inlay) were sectioned with a diamond saw in a BOL direc-
tion from the cusp tips. The sectioned surface was examined
294 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295
to measure the resin cement thickness at the buccal-lingual
tooth restoration interfaces. The four measuring points for
each sample are as follows; at two occlusal margins and at
the middle portions of the cement space of two buccal lin-
gual walls. The average cement gap was 117.4 (±53.8) ␮m.
The cusp flexure might depend on the fitness of inlay; the
more fit the inlay was, the lower cusp flexure would be
measured.
The composite (Z250) and resin cement (Rely X ARC) poly-
merization shrinkages were measured in a supplementary
experiment, using the ‘bonded disc method’ developed by
Watts and Cash [34]. The polymerization shrinkage of Rely X
ARC resin cement (4.28%) was 1.91 times higher than that of Z-
250 composite (2.24%). This may be because resin cement has
a lower filler volume than composite, giving it greater flow and
low film thickness.
After sitting the composite inlay, but before initiation of
light curing, a slight progressive cusp deflection was observed
(Fig. 3b). This may be because the resin cement is dual-cured,
and polymerization shrinkage occurred during the chemical
curing initiated after mixing. In addition, a transient, slight
decrease in cusp deflection was observed after light curing
began that was immediately followed by increasing again after
completion of light exposure. It seems that this is due to the
thermal expansion of the resin inlay body by the heat of the
curing light, which could transiently offset to a certain extent
the polymerization shrinkage of the resin cement [13,35,36].
In a supplementary experiment, it was observed that the tem-
perature at the tip of the light guide increased from 25 to 43 ◦ C
despite the LED light source, when measured using a temper-
ature sensor.
Hydration of the teeth may influence the amount of cuspal
deformation. Suliman et al. [12] reported that the hydrated
teeth had less cuspal deflections than did with dry teeth
because of possible hygroscopic expansion of the composite
or marginal gap formation. It is well known that the hydra-
tion of the specimen used in the lab is not perfect. In this
study, no special equipment such as an intra pulp chamber
irrigation device through apical foramen was used due to dif-
ficulty in setting up the measuring instrument, however, the
storage of the teeth, preparation of the cavity and measure-
ment were undertaken with a special care to prevent from
dehydration.
The cusp deflections due to the base under the composite,
the type of base, the type of composite or resin cement, and
mode of light curing are subjects for future study. In addition to
MOD cavity preparations, the cusp deflection during restora-
tion of Class V lesions with composite, in which postoperative
sensitivity is often reported, also requires further study.
In summary, the present study showed that the cusp deflec-
tion increased with increasing cavity dimension and C-factor,
and use of an incremental filling technique or an indirect com-
posite inlay restoration could reduce the cuspal strain.
Acknowledgement
This investigation was supported in part by a grant (03-PJ1-
PG1-CH09-0001) from the Ministry of Health and Welfare of
Korea.
references
[1] de Gee AF, Feilzer AJ, Davidson CL. True linear
polymerization shrinkage of unfilled resins and composites
determined with a linometer. Dent Mater 1993;9:11–4.
[2] Song YX, Inoue K. Linear shrinkage of photo-activated
composite resins during setting. J Oral Rehabil
2001;28:335–41.
[3] Lee IB, Cho BH, Son HH, Um CM. A new method to measure
the polymerization shrinkage kinetics of light cured
composites. J Oral Rehabil 2005;32:304–14.
[4] Rosin M, Urban AD, Gartner C, Bernhardt O, Splieth C, Meyer
G. Polymerization shrinkage-strain and micro-leakage in
dentin-bordered cavities of chemically and light-cured
restorative materials. Dent Mater 2002;18:521–8.
[5] Goldman M. Polymerization shrinkage of resin-based
restorative materials. Aust Dent J 1983;28:156–61.
[6] Rees JS, Jacobsen PH. The polymerization shrinkage of
composite resins. Dent Mater 1989;5:41–4.
[7] Lai JH, Johnson AE. Measuring polymerization shrinkage of
photo-activated restorative materials by a water-filled
dilatometer. Dent Mater 1993;9:139–43.
[8] Opdam NJ, Roeters FJ, Feilzer AJ, Verdonschot EH. Marginal
integrity and postoperative sensitivity in Class 2 resin
composite restorations in vivo. J Dent 1998;26:555–62.
[9] Carvalho RM, Pereira JC, Yoshiyama M, Pashley DH. A review
of polymerization contraction: the influence of stress
development versus stress relief. Oper Dent 1996;21:17–24.
[10] Segura A, Donly KJ. In vitro posterior composite
polymerization recovery following hygroscopic expansion. J
Oral Rehabil 1993;20:495–9.
[11] Alomari QD, Reinhardt JW, Boyer DB. Effect of liners on cusp
deflection and gap formation in composite restorations.
Oper Dent 2001;26:406–11.
[12] Suliman AH, Boyer DB, Lakes RS. Polymerization shrinkage
of composite resins: comparison with tooth deformation. J
Prosthet Dent 1994;71:7–12.
[13] Meredith N, Setchell DJ. In vitro measurement of cuspal
strain and displacement in composite restored teeth. J Dent
1997;25:331–7.
[14] McCullock AJ, Smith BG. In vitro studies of cusp
reinforcement with adhesive restorative material. Br Dent J
1986;161:450–2.
[15] Suliman AA, Boyer DB, Lakes RS. Interferometric
measurements of cusp deformation of teeth restored with
composites. J Dent Res 1993;72:1532–6.
[16] Pearson GJ, Hegarty SM. Cusp movement of molar teeth with
composite filling materials in conventional and modified
MOD cavities. Br Dent J 1989;166:162–5.
[17] Jantarat J, Panitvisai P, Palamara JE, Messer HH. Comparison
of methods for measuring cuspal deformation in teeth. J
Dent 2001;29:75–82.
[18] Hood JA. Biomechanics of the intact, prepared and restored
tooth: some clinical implications. Int Dent J 1991;41:25–32.
[19] Davidson CL, de Gee AJ. Relaxation of polymerization
contraction stresses by flow in dental composites. J Dent Res
1984;63:146–8.
[20] Unterbrink GL, Liebenberg WH. Flowable resin composites
as “filled adhesives”: literature review and clinical
recommendations. Quintessence Int 1999;30:249–57.
[21] Versluis A, Douglas WH, Cross M, Sakaguchi RL. Does an
incremental filling technique reduce polymerization
shrinkage stresses? J Dent Res 1996;75:871–8.
[22] Abbas G, Fleming GJ, Harrington E, Shortall AC, Burke FJ.
Cuspal movement and microleakage in premolar teeth
restored with a packable composite cured in bulk or in
increments. J Dent 2003;31:437–44.
 d e n t a l m a t e r i a l s 2 3 ( 2 0 0 7 ) 288–295
[23] Rees JS, Jagger DC, Williams DR, Brown G, Duguid W. A
reappraisal of the incremental packing technique for light
cured composite resins. J Oral Rehabil 2004;31:81–4.
[24] Rees JS, Jacobsen PH. Stresses generated by luting resins
during cementation of composite and ceramic inlays. J Oral
Rehabil 1992;19:115–22.
[25] Ericson D, Paulsson L, Sowiak H, Derand T. Reduction of cusp
deflection resulting from composite polymerization
shrinkage, using a light-transmitting cone. Scand J Dent Res
1994;102:244–8.
[26] Davidson CL, de Gee AJ. Light-curing units, polymerization,
and clinical implications. J Adhes Dent 2000;2:167–73.
[27] Feilzer AJ, De Gee AJ, Davidson CL. Increased wall-to-wall
curing contraction in thin bonded resin layers. J Dent Res
1989;68:48–50.
[28] Feilzer AJ, De Gee AJ, Davidson CL. Setting stress in
composite resin in relation to configuration of the
restoration. J Dent Res 1987;66:1636–9.
[29] Davidson CL, Feilzer AJ. Polymerization shrinkage and
polymerization shrinkage stress in polymer-based
restoratives. J Dent 1997;25:435–40.
295
[30] Davidson CL, Van Zeghbroeck L, Feilzer AJ. Destructive
stresses in adhesive luting cements. J Dent Res
1991;70:880–2.
[31] Craig RG, Powers JM. Restorative dental materials. 11th ed.
Mosby Inc.; 2002. p. 87–8.
[32] Blank JT. Scientifically based rationale and protocol for use
of modern indirect resin inlays and onlays. J Esthet Dent
2000;12:195–208.
[33] Alster D, Feilzer AJ, de Gee AJ, Davidson CL. Polymerization
contraction stress in thin resin composite layers as a
function of layer thickness. Dent Mater 1997;13:
146–50.
[34] Watts DC, Cash AJ. Determination of polymerization
shrinkage kinetics in visible-light-cured materials: methods
development. Dent Mater 1991;7:281–7.
[35] Sakaguchi RL, Versluis A, Douglas WH. Analysis of strain
gauge method for measurement of post-gel shrinkage in
resin composites. Dent Mater 1997;13:233–9.
[36] Feng L, Suh BI. The effect of curing modes on polymerization
contraction stress of a dual cured composite. J Biomed Mater
Res Part B: App Biomater, 2005 (published online).