Shoichi Ishigaki Tamaki Nakano Shinichi Yamada Takashi Nakamura Fumio Takashima

Biomechanical stress in bone surrounding an implant under simulated chewing

Authors afliation: Schoichi Ishigaki, Tamaki Nakano, Shinichi Yamada, Takashi Nakamura, Fumio Takashima, Osaka University Graduate School of Dentistry, Course for Integrated Oral Sciences and Stomatology, Division of Oromaxillofacial Regeneration, Department of Occlusion, TMD, and Advanced Prosthodontics Correspondence to: Shoichi Ishigaki, D.D.S., Ph.D. Department of Occlusion, TMD, and Advanced Prosthodontics Osaka University Graduate School of Dentistry 18, Yamadaoka, Suita Osaka 5650871 Japan Tel: 81 66879 2946 Fax: 81 68792947 e-mail: ishigaki/dent.osaka-u.ac.jp,

Key words: chewing function; nite element analysis; implant. Abstract: The concept of reducing nonaxial loading of dental implants has been widely regarded as the standard procedure. The aim of this study was to reveal the biomechanical stress distribution in supporting bone around an implant and a natural tooth under chewing function. Three-dimensional nite element models of the mandibular rst molar and the titanium implant both with the mandible in the molar region were constructed. The directions of displacement constraints were determined according to the angles of the closing pathways of chopping type and grinding type chewing patterns. The tooth model showed smooth stress distribution in the supporting bone with low stress concentration around the neck of the tooth. The implant model showed stress concentration in the supporting bone around the neck of the implant, especially in the buccal area. The grinding type model of the implant showed higher tensile stress concentration than the chopping type model at the lingual neck of the implant. The results of this study suggested the importance of considering occlusion under chewing function for understanding the biomechanics of oral implants.

Date:

18 February 2002
To cite this article:

Ishigaki S, Nakano T, Yamada S, Nakamura T, Takashima F. Biomechanical stress in bone surrounding an implant under simulated chewing Clin. Oral Impl. Res, 14, 2003; 97102 Copyright C Blackwell Munksgaard 2003 ISSN 0905-7161

The osseointegrated implant-supported restoration has become one of the most effective prosthodontic modalities in complete and partially edentulous patients. Since the osseointegrated implant has direct bone contact, the occlusal force on the implant superstructure is directly transmitted to the surrounding bone structure. Overloading on the superstructure could lead to the loss of osseointegration, and it has been clinically proposed to avoid high stress concentration in the supporting bone (Lindquist et al. 1988). Therefore, it is important to reveal how the occlusion of an implant-supported restoration affects the stress distribution in the supporting bone (Weinberg & Kruger 1995). However, boneloading patterns in nite element models are very sensitive to the various parameters and the creation of patient-dependent -

nite element models is required (Van-Oosterwyck et al. 1998). Attempts to reduce nonaxial loading of dental implants are commonly advocated but cannot be justied from the available research, at least from the standpoint of implant and bone health. Based on the currently available scientic evidence, it cannot be ruled out as one gold standard for implantation (Taylor 2000). Various studies have been conducted to investigate the load in bone and implant interface. Among these, the material of the xture (Takuma et al. 1990), design of the xture (Siegele & Soltesz 1989), the material of the superstructure (Davis et al. 1988), stress-absorbing material (Holmes et al. 1994), cantilever length (van Zyl et al. 1995), and the mechanism of bone and implant interface (Riger et al. 1989) have been reported to af-

97

Ishigaki et al . Biomechanical stress under simulated chewing

fect stress distribution in the supporting bone. For biting forces on dental implants, the basic problem is to determine the in vivo loading components on implants in various situations (Brunski 1992). Since loading is increasingly believed to be a determining factor in the treatment outcome with oral implants, there is a need to expand the knowledge related to the biomechanics of oral implants and the inuencing factors (Duyck et al. 2000). However, there has been no study of the effect of the functional occlusion of an implantsupported restoration on stress distribution in the supporting bone. Since the direction of the occlusal force is the main variable in stress distribution, the occlusion of an implant-supported restoration should be evaluated under simulated chewing function. The aims of this study were to investigate the closing direction of the mandible during chewing movement and to reveal the differences in stress distribution in the supporting bone between an implant and a natural tooth under occlusal function.

Fig.1. The three-dimensional nite element model of a natural tooth.

Material and methods

Construction of the 3D-FE models

Three-dimensional nite element models were constructed for the mandibular rst molar and the titanium implant including the mandible in the molar region. The implant structure was screw type titanium, 10 mm in length and 4 mm in diameter. A ceramo-metal crown was placed as the superstructure of the implant. The model of the occlusal table of the maxillary rst molar was also constructed as an antagonistic tooth. The occlusal relationship of the maxillary and the mandibular occlusal surface consisted of tripodism at the intercuspal position. The models were xed at the base of the maxillary rst molar. In order to place food elements between the maxillary and the mandibular occlusal surface, the mandible was placed 0.5 mm below the intercuspal position. The three-dimensional nite element model of the natural tooth consisted of 2214 nodes and 1766 solid elements (Fig. 1). The model of the implant consisted of 2269 nodes and 1731 solid elements (Fig. 2). Element discretization was repeatedly examined and modied according to the distortion check in the software (ANSYS,

Fig.2. The three-dimensional nite element model of an implant

Swanson Analysis Systems, Inc., Houston, TX, USA) on a personal computer (TITAN, Kubota, Tokyo, Japan). These processes led to validity of the model, indicating that element renement was completed with acceptable element distortion. The models were thus found to be sufciently valid for stress analyses.
Establishment of the loading conditions for the analyses

Twenty adult volunteers with no missing teeth and no signs of temporomandibular disorders were selected. Among them, 10 subjects (nine males and one female, age

2332 years, mean age 25.6) showed chewing patterns of a chopping type (Ctype), while the other 10 subjects (nine male and one female, age 2430, mean age 26.0) showed chewing patterns of a grinding type (G-type). Chewing movement was recorded with a jaw-tracking device (Sirognathograph Analyzing System III, Todent, Tokyo, Japan). The test food was a piece of commercial chewing gum (Free Zone, Lotte, Tokyo, Japan). The average of 10 chewing cycles after 5 s of mastication on the left and right side were analyzed. The directions of the closing pathways of the man-

98 |

Clin. Oral Impl. Res. 14, 2003 / 97102

Ishigaki et al . Biomechanical stress under simulated chewing

Table 1. Mechanical properties of different components of the models for the analyses Elastic modulus (Mpa)
1.37 104 1.37 103 8.41 104 1.40 104 1.00 10 1.00 10 6.72 104 9.10 104 1.07 105 1.00 10

Component
Cortical bone Cancellous bone Enamel Dentin Pulp Periodontal membrane Porcelain Gold alloy Titanium Food

Poissons ratio
0.30 0.30 0.30 0.15 0.49 0.49 0.30 0.33 0.33 0.30

chewing (60.7) (t-test, P 0.001). The mean angle of the closing pathways in the sagittal plane was signicantly greater in C-type chewing (22.2) than G-type chewing (1.3) (t-test, P 0.001). The mean angle of closing pathways in the horizontal plane was signicantly smaller in C-type chewing (60.3) than G-type chewing (89.3) (t-test, P 0.001) (Fig. 3). These results revealed that the closing pathway of G-type chewing is deviated toward the chewing side and anteriorly compared to that of C-type chewing.
Stress distribution in the supporting bone

paratively high maximum principle stress (tensile stress) concentration at the lingual neck of the implant in the buccolingual direction. No remarkable stress concentration was found in the supporting bone below the neck of the implant (Fig. 5).
Stresses in food elements

dible during the chewing movements, both in C-type and G-type, were calculated as the angle between the closing pathways within 0.5 mm from centric occlusion. Directions of displacement constraints at the bottom of the mandible simulated the calculated directions of the closing pathways for C-type and G-type chewing. The magnitude of the chewing force was selected as 200 N. The sum of the nodal force on the mandibular occlusal surface was 200 N after displacement constraints were applied to the model. The models were restrained at the base of the maxillary rst molar to avoid sliding of the entire model. The mechanical properties of the components in these models were based on previous studies (Takuma et al. 1990) (Table 1).
Stress analyses

around the natural tooth and implant

The natural tooth model showed smooth stress distribution in the supporting bone in the mesiodistal direction. It also showed smooth stress distribution in the supporting bone in the buccolingual direction except at the furcated area. Minimum principle stress (compressive stress) was found at the furcated area. The G-type model showed a similar stress distribution to the C-type model (Fig. 4). The implant model showed comparatively high minimum principle stress (compressive stress) concentration in the supporting bone at the mesial and distal neck of the implant in the mesiodistal direction. It also showed high minimum principle stress (compressive stress) concentration at the buccal neck of the implant in the buccolingual direction. Furthermore, the G-type model showed com-

Both the natural tooth model and the implant model showed almost the same stress values in food elements. However, C-type and G-type models showed different stress values in food elements. The average of the minimum compressive stress was higher in the G-type model than in the C-type model, but standard deviation (sd), coefcient of veriation (c.v.) and minimum value of the minimum compressive stress were higher in the C-type model than in the G-type model. On the other hand, average, sd and maximum value of the maximum shear stress was higher in the G-type model than in the C-type model. There was no difference in c.v. of the maximum shear stress between the C-type and the Gtype models (Table 2).

Discussion
To investigate the mechanical stress in bone surrounding an implant, various studies have been conducted using photo elastic stress analysis (Deines et al. 1993; Waskewics et al. 1994), mathematical analytic method (Morgan & James 1995; Wein-

Maximum and minimum principle stresses were analyzed in the supporting bone around an implant and a natural tooth. In order to quantitatively evaluate the stresses in the supporting bone, it was divided into ve areas in buccolingual and mesiodistal directions, respectively. Minimum compressive stress and maximum shear stress in the food elements between the maxillary and the mandibular occlusal surface were also evaluated. The stresses in the food elements were also evaluated in order to evaluate chewing function.

Results
Directions of closing pathways of the mandible during chewing movement

The mean angle of the closing pathways in the frontal plane was signicantly smaller in C-type chewing (35.6) than G-type

Fig.3. The directions of closing pathways of the mandible during chewing movements both in chopping type and grinding type.

99 |

Clin. Oral Impl. Res. 14, 2003 / 97102

Ishigaki et al . Biomechanical stress under simulated chewing

Fig.4. Stress distribution in supporting bone around a natural tooth is displayed along mesiodistal section and buccolingual section. M: mesial, MA: apex, F: furcated, DA: apex, D: distal, B: buccal, A: apex, L: lingual, smax: maximum principal stress, smin: minimum principal stress.

Fig.5. Stress distribution in supporting bone around an implant is displayed in mesiodistal section and buccolingual section. See abbreviations in Fig. 4.

Table 2. Stress values in food elements of the natural tooth model and the implant model Natural tooth model Average SD C.V.
0.24 0.19 C.V.

Implant model Minimum value

4.69 4.52 Maximum value

Average
2.75 2.97 Average

SD
0.66 0.56 SD

C.V.
0.24 0.19 C.V.

Minimum value
4.70 4.52 Maximum value

Minimum compressive stress C-type 2.75 0.64 G-type 2.97 0.57 Average Maximum share stress C-type 0.94 G-type 1.10 SD

0.24 0.28

0.26 0.25

1.61 1.72

0.94 1.11

0.25 0.28

0.26 0.25

1.61 1.72

C- type: Chopping type. G- type: Grinding type (Unit: MPa).

berg & Kruger 1995), and nite element stress analysis (Riger et al. 1989; van Zyl et al. 1995). Among these methods, nite element stress analysis is practicable to quantify internal stress in the model. The load transfer from implants to surrounding bone depends on the type of loading, the boneimplant interface, the length and diameter of the implants, the implant shape, structure of the implant surface, the superstructure and the quality and quality of the surrounding bone (Duyck et al. 1997). Further, the point of application, direction and magnitude of loads, and mechanical properties may easily be varied to simulate various situations. These studies utilized force from a vertical, horizontal or oblique direction toward the occlusal surface as the loading condition, which was not equivalent to real loading during a chewing movement. However, there has been no study that has analyzed the loading condition under chewing function. Mastication induces vertical and transverse forces, which induce axial forces and bending moments and exert stress gradients in the implant as well as in the bone (Duyck et al. 1997). Therefore, in this study, the directions of the closing pathways of the mandible during chewing movement were investigated. During chewing movement, occlusal force is generated by many closing muscle activities and is transmitted to the periodontal tissue. It is obvious that the occlusion affects the periodontal tissue structure. Overloading on the occlusal surface of the natural tooth leads to trauma of the periodontal tissue, tooth mobility and supporting bone resorption. On the other hand, the major complications associated with dental implants have become restorative-related rather than surgery-related. Because implants lack the stress release associated with a periodontal ligament, impact loading to restorative materials and the crestal bone remain potentially more damaging with implant-supported restorations (Curtis et al. 2000). Overloading of the superstructure of an implant could also lead to the loss of osseointegration (Lindquist et al. 1988; Quirynen et al. 1992; Hoshaw et al. 1994). Since there is an intimate correlation between occlusion, periodontal tissue structure and the osseointegration, it is important to evaluate the occlusion and to control the occlusal force on the natural tooth and implant. The natural tooth model showed

100 |

Clin. Oral Impl. Res. 14, 2003 / 97102

Ishigaki et al . Biomechanical stress under simulated chewing

smooth stress distribution in the supporting bone in both the mesiodistal and the buccolingual directions except at the furcated area. Conversely, the implant model showed stress concentration in the supporting bone around the neck of the implant in both the mesiodistal and the buccolingual directions. The absence of a periodontal membrane and the narrow neck of the implant could lead to this result. In particular, high minimum principle stress (compressive stress) was produced in the buccal area. Although the main component of the load was in a vertical direction during the chewing movement, the slight lateral component of the load in the buccolingual direction causes a stress concentration in the supporting bone at the buccal neck of the implant. From these results, it was suggested that the occlusion plays an important role in the stress distribution. Both the natural tooth model and the implant model have almost the same stress values in food elements. This result was consistent with previous studies, where there was almost the same chewing function between the subjects with natural dentition and implant-supported restorations (Haraldson & Carlsson 1979; Haraldson 1983; Strub et al. 1994). The C-type and G-type models showed different stress values in food elements. The C-type model, based on chopping movement, showed large compressive function in a part of food elements. The G-type model, based on grinding movement, showed large shearing function produced in the whole food elements. In this study, it was concluded that: O by determining the directions of displacement constraints using the angles of closing pathways during chewing movement, stress analyses could be executed under the simulated occlusal function; O the natural tooth model showed smooth stress distribution in supporting bone, while the implant model showed high stress concentration in supporting bone around the neck of the implant, especially at the buccal area; O the C-type and G-type models showed different stress values in food elements.

Although it was recommended that the posterior cusp inclination should be markedly reduced to avoid harmful lateral force (Weinberg 1998), the results of this study suggest that the occlusal form should be used for implant-supported restorations so as to correspond to the chewing pattern of the individual patient.

Implantatmodell des Knirschers zeigte sowohl lingual, wie auch besonders ausgeprgt buccal am Implantathals ein strkeres Dehnkrftemuster als bei einem Hackbiss. Das Implantatmodell des Knirschers zeigte lingual am Implantathals ein strkeres Dehnkrftemuster als bei einem Hackbiss. Die Resultate dieser Studie unterstreichen die Wichtigkeit der Okklusion whrend des Kauvorgangs fr das Verstndnis der Biomechanik von Zahnimplantaten.

sume Re
duction de la charge non-axiale des imLe concept de re te accepte comme un proplants dentaires a longtemps e tude a e te de valuer la cessus standard. Le but de cette e partition du stress bio-me canique au niveau de los aure tour dun implant et dune dent naturelle sous fonction `les nis en trois dimensions de de mastication. Des mode `re molaire infe rieure et dun implant en titane la premie te construits tous deux dans la re gion molaire de la ont e taient de termandibule . Les directions des contraintes e es en accord avec les angles des chemins de fermetumine `le dentaire re du type mastication et bruxisme Le mode partition de stress moelleuse dans los de montrait une re support avec moins de concentration de stress autour du `le dimplant montrait une collet de la dent. Le mode cer te concentration en stress dans los lentourant du co cialement au niveau de la zone vestibulaire. Le vical, spe `le du type bruxisme de limplant montrait une mode concentration de stress plus importante quau niveau du `le de mastication au niveau cervico-lingual de limmode cialement dans la zone vestibulaire. Le mode `le plant, spe du bruxisme de limplant montrait une concentration en `le de mastication au stress plus importante que le mode sultats de cette niveau cervico-lingual de limplant. Les re tude demontrent limportance de locclusion pour come canique des implants buccaux. prendre laction biome

Resumen
El concepto de reducir la carga no axial de los implantes dentales ha sido ampliamente considerado como el procendar. La intencio n de este estudio fue revedimiento esta nica del estre n biomeca s en el hueso de lar la distribucio soporte alrededor de un implante y un diente natural bajo funciones masticatorias. Se construyeron tres modelos de elementos dimensionales nitos del primer molar infebula en rior y del implante de titanio ambos en la mand n molar. Las direcciones de las fuerzas de desplazala regio ngulos de ciemiento se determinaron de acuerdo a los a n de tipo vertical y lateral. rre los patrones de masticacio una distribucio n suave del El modelo de diente mostro s en el hueso de soporte con una baja concentracio n estre s alrededor del cuello del diente. El modelo de del estre una concentracio n del estre s en el hueso implante mostro de soporte alrededor del cuello del implante especialmenrea bucal. El modelo de masticacio n de tipo latete en el a una concentracio n de estre s tensional mas alta ral mostro n de tipo vertical en el cuello que el modelo de masticacio rea bucal. Los lingual del implante especialmente en el a resultados de este estudio sugirieron la importancia de n bajo funciones de masticacio n para considerar la oclusio nica de los implantes orales. entender la biomeca

Zusammenfassung
Durch supponierte Kaubewegungen ausgelste biomechanische Krfte im Knochen um ein Implantat. Allgemein hat sich der Grundsatz durchgesetzt, dass nichtaxiale Krfte bei Zahnimplantaten wenn immer mglich zu reduzieren oder zu vermeiden sind. Das Ziel dieser Studie war es, die Verteilung der biomechanischen Krfte im Knochen um Implantate und natrliche Zhne whrend dem Kauvorgang zu untersuchen. Man baute ein dreidimensionales Modell des unteren ersten Molaren und eines Titanimplantates in der Unterkiefermolarenregion auf. Bei einem Hackbiss und bei einem Knirscher wurden die Winkel zwischen der Richtungen der aufgezwungenen Bewegungen und der Kieferschlussbewegung aufgezeichnet. Das Zahnmodell zeigte ein ausgeglichenes Krftemuster im umgebenden Knochen, mit geringer Krftekonzentration um den Zahnhals. Das Implantatmodell zeigte am Zahnhals, und dort vor allem buccal, eine Krftekonzentration im sttzenden Knochen. Das

101 |

Clin. Oral Impl. Res. 14, 2003 / 97102

Ishigaki et al . Biomechanical stress under simulated chewing

References
Brunski, J.B. (1992) Biomechanical factors affecting the bonedental implant interface. Clinical Materials 10: 201. Curtis, D.A., Sharma, A., Finzen, F.C. & Kao, R.T. (2000) Occlusal considerations for implant restorations in the partially edentulous patient. Journal of the California Dental Association 28: 779. Davis, D.M., Rimrott, R. & Zarb, G.A. (1988) Studies on frameworks for osseointegrated prostheses: part 2. The effect of adding acrylic resin or porcelain to form the occlusal superstructure. International Journal of Oral and Maxillofacial Implants 3: 280. Deines, D.N., Eich, J.D., Cobb, C.M., Bowles, C.Q. & Johnson, C.M. (1993) Photoelastic stress analysis of natural teeth and three osseointegrated implant designs. International Journal of Periodontics and Restorative Dentistry 13: 549. Duyck, J., Naert, I.E., Van-Oosterwyck, H., Van-der-Sloten, J., De-Cooman, M., Lievens, S. & Puers, B. (1997) Biomechanics of oral implants: a review of the literature. Technology and Health Care 5: 273. Duyck, J., Van-Oosterwyck, H., Vander-Sloten, J., De-Cooman, M., Puers, R. & Naert, I. (2000) Magnitude and distribution of occlusal forces on oral implants supporting xed prostheses: an in vivo study. Clinical Oral Implants Research 11: 465475. Haraldson, T. (1983) Comparisons of chewing patterns in patients with bridges supported on osseointegrated implants and subjects with natural dentitions. Acta Odontologica Scandinavica 41: 208. Haraldson, T. & Carlsson, G.E. (1979) Chewing efciency in patients with osseointegrated oral implant bridges. Swedish Dental Journal 3: 191. Holmes, D.C., Haganman, C.R. & Aquilino, S.A. (1994) Deection of superstructure and stress concentration in the IMZ implant system. International Journal of Prosthodontics 7: 246. Hoshaw, S.J., Brunski, J.B. & Cochran, G.V.B. (1994) Mechanical loading of Branemark implants affects interfacial bone modeling and remodeling. International Journal of Oral and Maxillofacial Implants 9: 360. Lindquist, L.W., Rockler, B. & Carlsson, G.E. (1988) Bone resorption around xtures in edentulous patients treated with mandibular xed tissue-integrated prostheses. Journal of Prosthetic Dentistry 59: 63. Morgan, M.J. & James, D.F. (1995) Force and moment distributions among osseointegrated dental implant. Journal of Biomechanics 28: 1109. Quirynen, M., Naert, I. & van Steenberghe, D. (1992) Fixture design and overload inuence marginal bone loss and xture success in the Branemark system. Clinical Oral Implants Research 3: 111. Riger, M.R., Adamus, W.L.C., Kinzel, G.L. & Brose, M.O. (1989) Finite element analysis of bone-adapted and bone-bonded endosseous implants. Journal of Prosthetic Dentistry 62: 440. Siegele, D. & Soltesz, U. (1989) Numerical investigations of the inuence of implant shape on stress distribution in the jaw bone. International Journal of Oral and Maxillofacial Implants 4: 340. Strub, J.R., Mylonas, T., Beyer, T. & Weingart, D. (1994) Functional state of edentulous patients with implantsupported xed prostheses and implant-retained overdentures: Preliminary results. International Journal of Oral and Maxillofacial Implants 9: 521. Takuma, M., Tsutsumi, S., Kurokawa, F., Takashima, F., Miyauchi, S. & Maruyama, T. (1990) The inuence of materials difference on stress distribution and bone remodeling around alumina and titanium dental implants. Journal of Osaka University Dental School 30: 96. Taylor, T.D. (2000) Research directions in implant prosthodontics. International Journal of Prosthodontics 13: 271. Van-Oosterwyck, H., Duyck, J., Vander-Sloten, J., Vander-Perre, G., De-Cooman, M., Lievens, S., Puers, R. & Naert, I. (1998) The inuence of bone mechanical properties and implant xation upon bone loading around oral implants. Clinical Oral Implants Research 9: 418. Waskewics, G.A., Ostowisk, J.S. & Parks, V.J. (1994) Photoelastic analysis of stress distribution transmitted from a xed prosthesis attached to osseointegrated implants. International Journal of Oral and Maxillofacial Implants 9: 411. Weinberg, L.A. (1998) Reduction of implant loading with therapeutic 285. biomechanics. Implant Dentistry 7:

Weinberg, L.A. & Kruger, B. (1995) A comparison of implant/prosthesis loading with four clinical variables. International Journal of Prosthodontics 8: 433. van Zyl, P.P., Grundling, N.L., Jooste, C.H. & Terblanche, E. (1995) Three-dimensional nite element model of a human mandible incorporating six osseointegrated implants for stress analysis of mandibular cantilever prostheses. International Journal of Oral and Maxillofacial Implants 10: 57.

102 |

Clin. Oral Impl. Res. 14, 2003 / 97102