Comparative Stress Distribution of Implant-Retained Mandibular Ball-Supported and Bar-Supported Overlay Dentures: A Finite Element Analysis

Overlay dentures have been widely considered acceptable treatment options for the edentulous patient.1,2 Bone loss around dental implants and the subsequent implant failure have yet remained concerns to the clinician and the patient. Functional loading creates strain, influencing bone remodeling around implants. Therefore, proper biomechanical considerations are very important in the long-term survival of dental implants.3–5 

Different attachment types are currently being used to provide proper support for the implant-retained overlay dentures. Implant-retained ball-supported mandibular overlay dentures have shown satisfactory outcomes.6–10 Additionally, certain authors have reported superior stress distribution with the application of implant-retained mandibular ball-supported overlay dentures compared with bar-supported dentures.11–16 Others believe that ball attachments should be used in the short term as a transitional phase. Based on their studies, it may be better to use implant-retained mandibular bar-supported overlay dentures in the long term because bar attachments have shown more appropriate stress distribution.17 Different ideas have been discussed regarding “rigid vs flexible retention” through the literature, and the search for the better option is promising.11–16,18 Finite element analysis is a popular numerical method in stress analysis and has been used in dentistry for years.19,20 The finite element analysis method is based on a mathematical model which approximates the geometry and the loading conditions of the structures. Deformation and stress distributions in different loading conditions can be simulated with the aid of computers, and the most stressed areas can thus be evaluated.21 

Since the literature provided controversial findings regarding the efficacy and safety of ball/bar design, the present study was conducted to comparatively assess the stress distribution patterns of implant-retained mandibular ball-supported and bar-supported overlay dentures using finite element analysis.

The three-dimensional (3D) model of the mandible was prepared based on a computerized tomography (CT) scan projection of a patient. The internal border of the cortical bone was defined in the Photoshop CS4 environment (Adobe Systems Incorporated, San Jose, Calif) for several sequential sections. The final 3D model of the mandible was formed using SolidWorks (SolidWorks Corp, Concord, Mass) and NURBS (McNeel, Seattle, Wash) software programs. A bar and ball attachment were simulated using SolidWorks (SolidWorks Corp). Implant models were considered threadless to facilitate simulation. To compensate for this simplification, the implant-bone interface was considered to be static. Due to the resilience of the soft tissue, the contact area of the prosthesis and soft tissue was assumed to be frictionless. Different elements were first assembled in the forms of mandibular implant-retained ball-supported and bar-supported overlay dentures and then inserted into ANSYS Workbench platform 2.0 (ANSYS Inc, Osaka, Japan). Four-sided elements were used for a mesh (grid) generation. Smaller elements were generated where higher precision was needed. Static analysis of the models was also performed in the ANSYS (ANSYS Inc). Mechanical properties of the simulated materials are presented in the Table.

The loads were exerted based on the theory of Gysi22,23 (cusp contacts) on the balancing and anterior facets. Load A (60 N) was exerted during the simulated protrusive motion with correspondent cusps in contact, and load B (60 N) was exerted during the simulated laterotrusive motion with correspondent cusps in contact. After imposing the A and B forces on the prostheses (on second molar mesial, first molar mesial, and first premolar areas), their effect was assessed on 3 sites: (1) the implant-tissue interface, (2) the first implant-bone interface, and (3) the second implant-bone interface.

Larger extension of mucosal strain was observed especially in the distal areas of the abutments. The main strain-bearing areas were, in decreasing order, buccal shelf, peri-implant soft tissue, and the distance between the implant and the buccal shelf (Figure 1a).

Two areas on the buccal shelf, two areas around the implants, and one area on the working side were the main strain-bearing areas. The strain-bearing area on the working side was larger than that of the nonworking side. Strain was also detected on the buccal aspect of the ridge on the working side. Higher strain levels were found around the implants, especially around the cervical area of the working side implant (Figure 1b).

On the working side, the maximum and the minimum strain were 1664 με and 2 με, respectively. The highest strain levels were observed in the cervical area. The strain was reduced at the fixture apex (Figure 2a and b). On the nonworking side, the maximum and the minimum strain were 2453 με and 2 με, respectively. The strain concentration was evident in the cervical and the apical areas. The maximum strain level was observed in the junction of the vertical and the oblique components of the implant (Figure 3a and b).

On the working side, the maximum and the minimum strain were 2435 με and 2 με, respectively. The highest strain level was observed in the interface of the upper implant margin and bone. Strain was seen all around the implant collar (Figure 4a and b). On the nonworking side, the maximum and the minimum strain were 2435 με and 2 με, respectively. Maximum strain concentration was observed in the apical third. Also in the cervical area of the fixtures, strain accumulation within the mild overload limits was detected (Figure 5a and b).

The largest strain-bearing area was buccal shelf, and the main strain-bearing area was peri-implant tissues. The peri-implant strain-bearing area in the nonworking side was more extended (Figure 6a).

The maximum and the minimum strain were 1794 με and 1 με, respectively. Strain values did not exceed the pathologic threshold. Also, a narrow tape of strain was observed on the buccal wall in the vicinity of the working side fixture. The maximum strain-bearing area was on the buccal of the implants which lies within the adapted area of the Frost cycle (Figure 6b).

On the working side, the maximum and the minimum strain were 1557 με and 1 με, respectively. The maximum strain was accumulated in the apical area of the implant. This was indicative of vertical transmission of the load to the bone. The strain was mostly concentrated at the apical corners of the implant. This was higher in the working side compared with the nonworking side. Also, strain concentration in the nonworking side was mostly detected in the junction of the middle and apical thirds of the implant (Figure 2c and d). On the nonworking side, this strain was more significant on the lingual side. Strain was observed in the horizontal surface of the implant apex, especially around the implant margins (Figure 3c and d).

On the working side, the maximum and the minimum strains were 1794 με and 1 με, respectively. The maximum strain distribution was in the apical one third and more prominently in the lingual. On the lingual side of the junction between the apical and the middle one third of the fixture, a mild overload strain-bearing area was observed (Figure 4c and d). On the nonworking side, strain concentration was more evident in the apical area. Strain was concentrated in the lingual and distal of the junction of the apical and middle thirds. In addition, strain was concentrated at the apical surface of the fixture, which was indicative of vertical transmission of the load to the bone and its consequent strain (Figure 5c and d).

Implant-retained bar-supported overlay dentures are associated with superior strain distribution compared with the ball-supported designs. Strain was mostly detected in the apical of the fixtures and least in the cervical when bar design was used. The strain was mostly detected in the cervical of the working side implant. In the nonworking side, however, strain was higher in the cervical and lower in the apical compared with the working side implant. In posterior edentulous areas, a similar strain distribution pattern was detected, but still strain values were higher in the ball design.

Strain was mostly seen in the cervical and in the apical one third of the implants. In the bar design, strain was mostly accumulated in the middle, while middle area strain was considerably lower with the application of ball design. The tissue contact of prostheses in the ball design is higher than that of bar design. In both designs the strain accumulation was observed in the cervical of the implants. The strain values for both designs were less than the pathologic threshold proposed by Frost.24 The strain was distributed over a bigger volume in the ball design. The strain extension pattern was different in the cervical and the apical areas, and the least strain was observed in the middle one third. With the application of bar design, strain is mostly detected in the apical and middle one third. In the cervical one third (except for the crestal edge), however, strain-free areas are observed.

The dental implant literature confirms the adverse effects of improper bone loading.25–27 The present strain study aimed to assess the distribution pattern, to determine the main strain-bearing areas, and to find the superior support (ball or bar) in terms of strain distribution. In the present study, a complete mandibular model was created based on CT scan slices of a patient to further simulate natural conditions. The outcomes of the study were assessed based on the criteria described by Frost.25 The most important finding of the present study was the absence of the pathologic strain based on the Frost threshold in both designs.

The other important finding was the presence of the minimum strain tape that reflects the areas prone to atrophy due to lack of function. Of course, these areas may experience enough function under other loading directions, which will reduce the possibility of atrophic changes. In ball-supported prostheses, load was exerted laterally to the implant-bone interface; in bar-supported prostheses, load was mostly directed vertically to the interface.

Maximum strain was equal to 2440 με and 1557 με, respectively, for ball- and bar-supported designs in the cervical area of the fixtures. This pattern indicates a superior strain distribution pattern with the application of bar designs compared with that of ball design. The strain-bearing area associated with the ball-supported design was larger than that of the bar-supported design. It should be noted that clinical aspects such as the fitness of the bar-implant assembly or the simplicity of the process and the long-term results of bacteria accumulation at the junction of the ball/bar and implant are not of concern in the present study. It is then highly recommended to study clinically the superior support design for the implant-retained overlay dentures.

Clinically speaking, patient convenience, the screw loosening, non passive bar superstructure fit, and the consequent strains should also be considered in the comparison of bar and ball designs. Ideally, bar design is superior to ball design in terms of strain distribution. Again, the most important finding of the present study was the superior strain distribution within the bone surrounding the bar-supported implants.

The findings of the present study are consistent with the findings of Cekiç et al26 who comparatively studied the ball, bar, and cantilever designs on acrylic resin. They found that of the three designs, total strain was highest in the ball design. The authors suggested uneven force distribution in the ball design as the reason for the local strain accumulation.

Celik and Uludag27 assessed the photoelasticity of different bar and ball designs on 3 implants and found them to be all within the safe zone, except for the ball-supported, nonworking side implant. The forces on the working side in the present study were higher, which is due to the different designs and loads applied in the 2 studies. Barão et al28 reported lower maximum force with the application of ball compared with bar design. Although the difference may be due to the different loads applied, Barão et al28 applied vertical load to the central incisor area.

The findings of the present study are not supported by those of the study of Daas et al.29 In their study, the ball-supported implants were associated with higher strain in the soft tissue and lower strain values in the implants. The difference is thought to be due to the greater displacement of the Dalbo Plus design compared with that of the elastic gasket of the present study.

Within the limitations of the present study, the patterns of stress distribution of implant-retained bar-supported and ball-supported overlay dentures were similar. Compared with the ball design, the bar design resulted in smaller strain magnitudes both in laterotrusive and protrusive motions, distributed over smaller areas. Therefore, the bar design might be considered superior. However, bearing in mind that no areas of pathologic strain were observed over both designs, as well as the fact that physiologic strains might be a stimulant for bone remodeling, the ball design with a greater magnitude and distribution of nonpathologic strains could be advantageous as well. However, small areas of borderline strains observed over the ball design should be considered so as not to utilize such a design in a patient with excessive masticatory forces.

3D

three-dimensional

CT

computerized tomography

The authors would like to thank Farzan Clinical Research Institute for technical assistance.