Tables 5 and 6 show the maximum pressure on the mucosa and the contact area between the denture and mucosa respectively。 Under all three loading conditions, the maxi- mum pressure on the mucosa in model B was higher than in the other models, especially under VI。 The peak pressure in model B was observed under VI; it was approximately two times   as   high   as   in   the   other   three   models,   and was

concentrated between the labial side of the anterior alveolar ridge and the denture。

Under VM and IM, the contact area between the denture and mucosa was larger than that under VI。 Under VM, the contact area between the denture and mucosa in model A was about 75% of the area in the other models。 Under VI, the contact between the denture and mucosa mainly took place on the labial side of the anterior alveolar ridge for models B–D, whereas for model A it was concentrated on the left side of the whole alveolar ridge (Fig。 6)。 The contact area in model A was

Fig。 6 – Distribution of contact area between the denture and mucosa under a VI load。 (A–D) represent models A–D, respectively。 The cold tone represents the area where contact with the denture was close and tight, whereas the warm tone indicates the area where the denture tilted and separated from the mucosa。

about three times as large as in the other three models  under

VI。 This illustrates that, when functioning with the anterior teeth, the single-implant-retained overdenture rotated over the implant from one side to the   other。

4。 Discussion

The FE models used in the present study allows representation of a more detailed and complex geometry。 However, the inherent limitations of the FEA with regards to strain distribution should always be taken into consideration。22,23 The models used deviated in many aspects from a clinical situation。 The structures in the models were all assumed to be homogeneous, isotropic and linearly elastic。 However, it is well documented that the cortical bone of the jaw is transversely isotropic and inhomogeneous。 In addition, a 100% implant/bone interface was established, which does not match clinical situations。 Thus, the results of FEA of a problem like this should be interpreted with some care。 The absolute values of the different strains obtained in this study are of minor interest。 What are of interest are the relative values of the different strains for the different implant overdenture designs。 Therefore, the results we obtained should be considered as a reference to choose between different over- denture designs in the clinical treatment。 Prospective clinical studies are required to verify the results。

In previous studies, the interface between the denture and the mucosa was assumed to be fixed to facilitate modelling and calculations。28,29 However, our study assumed that sliding friction existed between the denture and the mucosa。 Our model of overdenture could therefore rotate and slide on the mucosa in various directions when functioning and so could more accurately simulate actual denture movement in daily use。 In addition, we assumed in the present study that, with regard to posterior loads, the opposite side would show the same mechanical behaviour as the loaded  side。

The results from our study indicated that in all models, maximum equivalent strains in peri-implant bone under all three loading conditions were below 2500 me, and were therefore lower than the  physiological  tolerance  threshold of bone。21 This findings agrees with previous clinical studies that showed no significant difference in peri-implant bone resorption between single-, two- and four-implant-retained/ supported overdentures, indicating that the strain in the bone around implants was within the physiological threshold, having little effect on the implant survival    rate。11,16,30