Dental Research Effect of restorative materials on cuspal flexure John MedigeVYuru Deng**/XJnyi Yu***/Elaine L. Davis****/Robert B. Joynt**** The purposes of this .study were {1) to establish a methodology for determining surface strains in two locations ofthc same tooth under intact, prepared, and restored conditions and(l) to compare the effects on stiffness of different restorative materials in a tooth subjected to cuspal loading. Two linear strain gauges were tnounted on each of 30 extracted ma.\iltary premolar teeth. Teeth were mounted in poly(methyl methacrylate) resin and randomly ussigned to one of three study groups according to the restorative material and application technique to be used. Statistical analysis indicated a statistically significant interaction between restorative material and tooth condition at both proximal and buccal sites and a statistically significant difference in stiffness between teeth restored with Tenure/Marathon V and those restored with either amalgam or Seotchbond 2/P-50 at the proximal site. Results suggest that the methods employed provide a useful, nondestruetive means of testing the same tooth under various conditions. (Quintessence inl 1995:26:571-576.) Introduction Preparation reduces the stmctnral integrity of teeth. Because resin composite can be bonded to enamel with the acid-etching technique, a more conservative cavity preparation is possible than for amalgam restorations. The use of dentinal bonding systems can fiirther expand the area of attachment between tooth structure and resin composite, increasing both mechanical stiffness and resistance to mieroleakage. * Associate Professor, Deparlmcnl of Mechanical and Aerospace Engineering, State Universily of New York at Buffalo, School of Engineering and Applied Sciences. Buffalo, New Yoflt '* Graduate Student, Department of Biomaterials, Stale University of New York at Buffalo, School of Dental Medicine, Buffalo, New York, **' Assistant Professor, Department of Restorative Dentistrj', State University of New York at Buffalo, School of Dental Medicine, Buffalo, New York, **** Associate Professor. Depanment of Restorative Dentistry, State University of New York at Buffalo, School of Dental Medicine, Buffalo New York. Reprint requests: Dr E. L, Davis, Associate Professor, Departmenl of Restorative Dentislry, State University of New York at Buffalo, Schoolof Denial Medicine, 335 Squire Hall, Soulh Campus, Buffalo, New York Í42I4, The use of resin composites to restore posterior teeth has been a controversial subject since resin composites came into widespread use in the late 1960s, Studies investigating the purported reinforce- ment effect of resin composites on weakened tooth structure have produced varied results.'"" Fracture tests by Gelb et al' indicated that resin- bonded restorations restore tooth strength lost after cavity preparation, Morin et aF found that tooth rigidlt>' is increased when an acid-etching technique is used to treat enamel before placement of resin composites. Eakle* found that the use of dentinal bonding agents further increases resistance to fracmre of teeth restored with resin. In most previous studies of cuspal flexure and fracture, a load was applied in an axial direction through a hard spherical or cylindrical testing device to the occlusal surface of the tooth. This procedure results in high stress concentrations on the triatigular ridges of the facial and lingual cusp inclines, rather than distribution of stress over the occlusal surface. The purposes of the present study were (I) to establish a methodology for determining the defor- mation of teeth caused by cuspai loading, using a method that distributes force over a substantial portion Ouintessejje*- 571
Effect of restorative materials on cuspal flexure
John MedigeVYuru Deng**/XJnyi Yu***/Elaine L. Davis****/Robert B.
Joynt****
The purposes of this .study were {1) to establish a methodology for
determining surface strains in two locations ofthc same tooth under
intact, prepared, and restored conditions and(l) to compare the
effects on stiffness of different restorative materials in a tooth
subjected to cuspal loading. Two linear strain gauges were tnounted
on each of 30 extracted ma.\iltary premolar teeth. Teeth were
mounted in poly(methyl methacrylate) resin and randomly ussigned to
one of three study groups according to the restorative material and
application technique to be used. Statistical analysis indicated a
statistically significant interaction between restorative material
and tooth condition at both proximal and buccal sites and a
statistically significant difference in stiffness between teeth
restored with Tenure/Marathon V and those restored with either
amalgam or Seotchbond 2/P-50 at the proximal site. Results suggest
that the methods employed provide a useful, nondestruetive means of
testing the same tooth under various conditions. (Quintessence inl
1995:26:571-576.)
Introduction
Preparation reduces the stmctnral integrity of teeth. Because resin
composite can be bonded to enamel with the acid-etching technique,
a more conservative cavity preparation is possible than for amalgam
restorations. The use of dentinal bonding systems can fiirther
expand the area of attachment between tooth structure and resin
composite, increasing both mechanical stiffness and resistance to
mieroleakage.
* Associate Professor, Deparlmcnl of Mechanical and Aerospace
Engineering, State Universily of New York at Buffalo, School
of
Engineering and Applied Sciences. Buffalo, New Yoflt
' * Graduate Student, Department of Biomaterials, Stale University
of New York at Buffalo, School of Dental Medicine, Buffalo, New
York,
**' Assistant Professor, Department of Restorative Dentistrj',
State University of New York at Buffalo, School of Dental Medicine,
Buffalo, New York,
**** Associate Professor. Depanment of Restorative Dentistry, State
University of New York at Buffalo, School of Dental Medicine,
Buffalo New York.
Reprint requests: Dr E. L, Davis, Associate Professor, Departmenl
of Restorative Dentislry, State University of New York at Buffalo,
Schoolof Denial Medicine, 335 Squire Hall, Soulh Campus, Buffalo,
New York Í42I4,
The use of resin composites to restore posterior teeth has been a
controversial subject since resin composites came into widespread
use in the late 1960s, Studies investigating the purported
reinforce- ment effect of resin composites on weakened tooth
structure have produced varied results.'""
Fracture tests by Gelb et al' indicated that resin- bonded
restorations restore tooth strength lost after cavity preparation,
Morin et aF found that tooth rigidlt>' is increased when an
acid-etching technique is used to treat enamel before placement of
resin composites. Eakle* found that the use of dentinal bonding
agents further increases resistance to fracmre of teeth restored
with resin.
In most previous studies of cuspal flexure and fracture, a load was
applied in an axial direction through a hard spherical or
cylindrical testing device to the occlusal surface of the tooth.
This procedure results in high stress concentrations on the
triatigular ridges of the facial and lingual cusp inclines, rather
than distribution of stress over the occlusal surface.
The purposes of the present study were (I) to establish a
methodology for determining the defor- mation of teeth caused by
cuspai loading, using a method that distributes force over a
substantial portion
Ouintessejje*- 571
Dental Research
Fig 1 Loading device against cusloni-made casting.
ofthe occlusal surface and (2) to compare the pattern of structural
deformation under load for intact, pre- pared (with
mesio-occlusodistal [MOD] prepara- tions), and restored teeth. A
maximal load of 300 N was used, which is within the range of normal
chewing forces.''^ Restorations were either bonded (resin
composite) or unbonded (amalgam).
Method and materials
Extracted maxillary premolar teeth were collected and placed in a
i% hydrogen peroxide solution imme- diately following extraction.
Teeth were examined visually with the aid of a transilluminating
fiberoptic light. Only those teeth with no visible defects were
retained. Maximal mesiodistal and buccolingual di- mensions were
measured with a Boley gauge (Buffalo Dental). These two dimensions
were totalled, and only those teeth within a specified range ( 16.6
to 19.3 mm) were included, to minimize differences in tooth
size.
Thirty teeth were selected, cleaned with pumice, scaled, and placed
in deionized water They were removed from water only long enough to
complete necessary procedures. Tiie teeth were randomly as- signed
to one of three groups, according to the restorative materials to
be used: dental amalgam or one of two posterior resin composite
restorative systems.
The apex of each tooth was centered on the base portion oía
two-piece break-apart form (SampMCup, Buehler), so that the long
axis of the tooth was perpendicular to the plane of the base. The
root portion of the tooth was then embedded in poly- ( methyl
methacrylate) (PMMA) tray resin (Tramix. Stratford-Cookson) to a
point approximately 2 mm below the cementoenamel junction (CEJ), to
approxi- mate the height of healthy alveolar bone. The base of each
mounted specimen was trimmed to expose a cross section of the root,
reducing the root length by approximately 3 mm. The exposed root
allowed transmission of applied force entirely through tooth
structure and prevented subsidence of tiie tooth in the resin
during testing.
A custom-designed loading head was fabricated for testing
specimens. The loading device was beveled based on the calculation
of average occlusal cuspal inclines for nine previously selected
maxillary pre- molars, 37 degrees on the buccal cuspal inciine and
33 degrees on the lingual cuspal inciine.
A casting, which adapted to both the cuspal inclines ofthe occiusai
surface and the beveled faces ofthe loading device, was made for
each specimen to distribute the load over the outer parts ofthe
occlusal surface (Fig 1). A strip of 28-gauge relief wax was
applied to the central portion ofthe occlusal surface of each
specimen. An impression was then made of each mounted specimen and
poured up with improved dental stone. The layer of wax provided a
spacer to prevent contact of the finished casting with the
restoration. The casts were removed from the impres- sion, trimmed,
and allowed to set for at least 24 hours.
The loading head was mounted in a surveyor and positioned so that a
wax pattern could be fabricated. Wax was applied to the occlusal
surface ofthe cast and extended just over the buccal and litigual
cusps. The mounted loading head was warmed and used to form the
upper surface ofthe wax pattern. The thickness of the wax pattern
was kept to about 1 mm.
The wax pattern was sprued and invested with Beauti Cast investment
material (Whip Mix), placed in a water bath for 1 hour, and then
cast with Williams Technique Metal 35 (lvoclar). The sprue was
removed and the surface of the casting was finished and polished.
The casring was cut into two pieces along the central groove to
allow the buccal and lingual cusps to flex independently during the
testing procedure.
A custom resin composite matrix was fabricated on the proximal
surfaces of each intact tooth. Rosin
572 Quintessence International Volume 26, Number 8/1995
Dental Research
Fig 2 Proximai view of strain gauges on buccai anc proximai
surfaces.
Fig 3 Dimensions of MOD cavity preparations.
A = 1/3 of B C = 1/3ofD
D
B
i A
composite (Opalux, ICI Pharmaceuticals) was adapt- ed to both
proximal surfaces and activated with a high-intensity light source
(Prisma Lite, LD Caulk) for 30 seconds, A thin layer of toil
{Matrix Strip. Den-Mat) was cemented to the inner surface ofthe
matrix to prevent adhesion ofthe restorative material,
A linear strain gauge (grid length 0,79 mm, nominal resistance 120
ohms; model EA-O6-O31DE-12O, Mea- surements Group) was mounted on
the buccal surface of each tooth in a vertical orientation. An
identical gauge was placed horizontally on enamel, at the proximal
CEJ (Fig 2), The backing ofthe proximal gauge was trimmed to
approximately 1 mm in width and placed immediately above the
proximal CEJ. Both gauges were primed with a catalyst before
attachment with an adhesive (M-bond 200, Measurements Group),
Bondable terminais were attached to the side of the PMMA base, with
the same adhesive, and wired to the gauges.
The installed gauges were then tested with a strain gauge
installation tester (Model 1300, Measurements Group) to assure
their functionality'. The strain gauge, solder, and connecting
wires were covered with adhesive (Mirro 3, Kerr/Sybron) and sealant
(737 RTV, Dow Corning) to prevent moisture contam- ination. After
the coatings were set. each specimen was placed in a container of
deionized water maintained at room temperature.
Three groups were established according to resto- rative material
to be used: amalgam. Scotchbond 2 with P-50 (3M Dental) and Tenure
with Marathon V (Den-Mat), Each mounted specimen was placed on the
lower platen of an axial testing machine (T22K, MTS}, The beveled
loading device was used to test each tooth at a crosshead speed of
1 mm/niin to a maximal load of 300 N,
Specimens in all three groups were prepared for MOD restorations
after they had been tested intact, Cavit>' dimensions are shown
in Fig 3, Hach cavity was cut with a No, 56 FG bur (SS White). All
specimens were then retested.
Specimens in the first group were restored with Valiant-Ph.D. (LD
Caulk}, a high-copper dental amalgam. The resin composite rnatrix
was secured whh a special retainer to protect the gauges. After
condensation, the matrix and retainer were removed and the
restoration was carved to normal anatomic form.
A 1-mm-wide occlusal marginal bevei was placed on the cavity
preparation of resin composite specimens, at an angle of
approximately 45 degrees to the unpre- pared surface. The enamel
portion of the cavity preparation and 1 mm beyond the cavosurface
margin were etched with 31% phosphoric acid for 30 sec- onds,'
Specimens were then rinsed with water and dried with a stream of
oil-free compressed air.
Quintes? \/niiimp 9P. Number 8/1995 573
Dental Research
Fig 4 MGan (± SE) proximal strain-force curves as a function oi
restorative material and toolh condition A higher mean (ie, closer
to íero¡ indicates greater stiffness.
Fig 5 Mean (+ SE) buccal strain-force curves as a function of
restorative material and tooth condition. A higher mean (ie, closer
to zero) indicates grealer stiffness.
Scotchbond 2 dentinal bonding agent was applied to the second group
of specimens. The matrix was placed and the preparation was
restored incrementaily with P-50 light-activated resin composite.'°
Each increment was light activated for 20 seconds, and the entire
restoration was exposed to the light source for an additional 60
seconds.
The Tenure dentinal bonding system was applied to the third group
of specimens. The matrix and retainer were placed and the MOD
cavity was restored with Marathon V resin composite, a light- and
chemically activated (ie. dual-cured) paste system, applied in a
syringe. The restorative material was light activated for 60
seconds.
The occlusal surface of each specimen was checked with articulating
paper to confirm that the casting did not touch the restoration.
Specimens were stored for at least 24 hours in 100% relative
humidity, to allow for complete set of the restorations, and then
retested.
Buccal and proximal strains, force, and loading head displacement
were recorded with a condilioner/ amplifier (2100 System,
Measurements Group) and a computerized data-acquisition system (IBM
PC XT, Data Translation DT 280ÍA) using a customized program and
commercial software (ASYST), Strain- force curves were plotted, and
linear regressions were obtained. Data were recorded for each tooth
under the
three conditions: intact, prepared but unrestored, and
restored.
Each test was repeated three times. The first run was used to
ensure proper seating and alignment. The second was used for data
acquisition. The third run was used for verification, Muitivariate
repeated-measures analysis of variance (ANOVA) and multiple-com-
parison procedures were used to analyze the data.
Results
Thiee specimens with MOD cavity preparations frac- tured during
testing and were eliminated from the analyses. Several other
specimens were eliminated because of failure ofthe strain gauge,
resulting in a total of 23 specimens (nine amalgam, seven
Scotchbond 2/P-5O and seven Tenure/Marathon V} for the proxi- mal
site and 23 specimens (eight amalgam, eight Scotchbond 2/P-50 and
seven Tenure/Marathon V) for the buccal site.
Means and standard errors ofthe strain-force slopes for the three
material groups under intact, prepared, and restored conditions are
shown in Figs 4 (proxi- mal) and 5 (buccal), A negative siope
indicates compressive (as opposed to tensile) strain.
Repeated-measures ANOVA indicated a significant interaction between
restorative material and tooth
574 Quintessence International Volume 26, Number
Dental Research
Source
Mean square = 3.70
• yl - total score, across conditions (used to lest nialerial main
efTect); y2 = difference score, restored minus prepared; y3 -
difference seore, intact minus prepared.
t Main effect for condition, P- .0005. Uniyariate results were
significant for both y2 (F,,20 = T.56; P- .012) and y3 (F, 20 '
31''O; P- 0002).
X Material by condition interaction, P ~ .020. Univariate results
were significant for y2 IF-, m» 7.63; P = .003) but nol for y3 (F,
,0 " 0.01; ^=.993).
Test of significance*
- F2,2o=0.22 Mean square = 1.53
* yl - tolal score, aeross conditions (used to test material rnain
effect]; y2 - difference seore, restored tninus prepared; y3 =
difference seore, intact minus prepared.
t Main effect for eondition.P^.0001, Univariate results wtre
significant for both y2(Fi,2o- 1 i. 16;/>= .0003) and y3
(Fi,i(,= 50.66;/>=.0001 !.
Í Material by condition interaction, P = .042. Univariate results
were significant for y2 ¡Fi 20 = ^•^'^- P = 026) but not for y3 IF2
lo = 0-33; P^.797).
condition (intact, prepared, or restored) at both proximal (F4 38 =
3.30; P- .02) (Table 1) and buccal sites (Fijg = 2.76, P - .04)
(Table 2). Univariate results indicated that this interaction was
significant for prepared versus restored, but not for intact versus
prepared, conditions at both sites.
Because the materiai-by-condition interaction was significant for
the prepared versus restored conditions but not intact versus
prepared conditions, follow-up multiple comparisons (Tukey HSD)
were made be- tween the prepared and restored data for each
restora- tive material. In addition, comparisons between mate-
rials at each location (proximal and buccai ) were made for both
the prepared and restored conditions. A significance level of .05
was used for ali comparisons. These tests indicated statistically
significant diiferences in slope between prepared and restored
conditions for Tenure/Marathon V only, at both the proximal and
buccal sites. In addition, statistically significant dif- ferences
under the restored condition were found between Tenure/Marathon V
and both amalgam and Scotchbond 2/P-50 at the proximal site.
These results indicated that teeth restored with Tenure/Marathon V
were stifier than teeth that were prepared but not restored.
Results did not provide staüstical evidence that teeth restored
with either Scotchbond 2/P-50 or amalgam were any stiffer than
prepared, unrestored teeth. At the proximal site, teeth restored
with Tenure/Marathon V were significantly
stiffer than were those restored with either amalgam or Scotchbond
2/P-50. Results did not provide statisticai evidence of a
difference in stiffness among groups at the buccai site.
Discussion
This study describes a nondestructive method of determining locai
tooth deformation, allowing re- peated ioading ofthe same tooth
under varied condi- tions. In addition, the use of individually
fabricated castings and the custom-made loading device provides a
means of appiying force to the same part of the occlusal surface in
all tests without ioading the restoration.
Specimens restored with amaigam showed little or no recovery of
tooth stiifness at either the proximai or buccal site. This was
anticipated, because amalgam does not bond to tooth structure.
Restoration with resin composite provided substantial recovery of
tooth stiffhess for Tenure/Marathon V specimens at both buccal and
proximal sites. For Scotchbond 2/P-5O specimens, resuits did not
provide evidence to suggest recovery of tooth stiffness at either
site.
Two factors, the bonding system and the resin composite, may have
contributed to the difference observed between the resin composite
groups. The two bonding systems differ in basic formulation and
curing method, and in the way in which the smear layer is
Quinte;to-nrnr Number 8/1995 575
Dental Research
treated. In the Tenure system, the conditioner (which contains 2,5%
nitric acid) removes the smear layer and opens dentinal tubule
orifices, ln Scotchbond 2, the primer (which contains 2,5% maleic
acid) dissolves the smear layer, and smear layer tags remain and
occlude the tubules. Tenure solutions A plus B
(N-tolyglycine-glycidyl methacr^'late plus pyromellitic acid
dimethacrylate) are chemically cured aqueous bonding materiais with
low film thickness, whereas Scotchbond 2 Is a light-cured,
resin-based bonding material with greater film thickness.
The resin composites in these two systems also difFer, P-SO is a
light-cured resin composite, while Marathon is a dual-cui-ed resin.
Determination of the potential contribution of each of these
factors—smear layer treatment, bonding system formulation and
curing method, and resin composite—requires turther study.
In this study, sutface strains were measured in two locations. It
would be imprudent to assume that these measurements alone indicate
the entire state of stress in the tooth and its likelihood to fail.
Strain is indicative of the local deformation only, and isolated
strain measurements are subject to misinterpretation, A tooth is a
complicated structure with complicated supporting structures and,
as such, is extremely difTicult to analyze. Detailed knowledge of
the geo- metry and of the mechanical properties of each part of the
tooth and its supporting structures is required. These properties
vary with location and direction. Thus, in the absence of more
accurate measurements and/or analyses, the strains at the location
where the stresses are likely to be relatively high might serve as
indicators of load severity. Furthermore, because each tooth is
tested under several conditions and is thus its own internal
control, the strains under the various conditions may reasonably be
construed to indicate the relative resistance to deformation under
those condi- tions.
Future research should focus on the effects of long-term cyclic
loading on tooth stiffness, because durability of restorative
systems is an important clinical consideration. Future studies
should also include a model that more closely simulates conditions
of the oral environment. In addition, although the present
investigation provided information regarding strain measurement on
the tooth surface, future work should determine internal strains
through the use of computational methods, such as finite-element
analysis.
References
1, Gelb MN, Barouch E. Simonsen RJ, Resistance to cusp fracture in
Class II prepared and restored premolars. J Prosthet Dent
I986;55;1S4-185.
2, Share J, Mishell Y, Nathanson D, Effect of restorative material
on resistance to fracture of tooth structure in vitro labstraet
622|, J Dent Res 19S2;61:247.
3, Joynt RB, Wieczkowski G Jr, laockowski K, Davis EL, Effects of
composite restorations on resistance to cuspal fracture m posterior
teeth, J Prosthet Dent 1987;57;431-435,
4, Stampalia LL, Nicholls JL Bnidvik JS, Jones DW, Fraclure
resistance of teeth with resin-bonded cestoratiotis, J Prosthet
Dent i986;55;694-69a,
5, Morin D, DeLong R, Douglas WH, Cusp reinforcement by the
acid-elch technique. J Denl Res 1984^63:1075-1078,
6, Eakle W, Fracture resistance of teetb restored with Class 11
bonded composite resin, J Dent Res I9K6,65:149-153,
7, Vïldmalm S-E, Ericsson SG. Maximal bite force with centric and
eccentric load. J Oral Rehabil 1982;9:445-450,
S. Gibbs CH, Mahan PE, Lundeen HC, Brehnan K, Walsh EK, Holbrook
WB. Occlusal forces during chewing and swallowing as measured by
sound transmission. J Prosthet Dent 1981:46:443-449.
9. Gilpatrick RO, Ross JA, Simonsen RJ, Resin to enamel bond
strengths with variable etching limes [abstract 13081, J Dent
Res
10. Weczkowski G Jr, Joynt RB, Kiockowski R, Davis EL, Effects of
incremental versus bulk fill techniqtie on resistance to cuspal
fracture of teeth restored with posterior composites, J Prosthet
Dent 1988;60:283-287, n
576 Quintessence International Volume 26, Number 8/ .935