PURPOSE: The aim of this study was to examine the bending moments and fracture patterns of different zirconia abutments with internal implant-abutment connections after static loading and to compare their bending moments to those of internally connected titanium abutments. Materials and METHODS: Three types of customized zirconia abutments (Straumann CARES abutments/Straumann BL implants [T1], Astra ZirDesign abutments/Astra Micro Thread OsseoSpeed implants [T2], Zirabut prototype abutments/Straumann SP implants [T3]) and one type of customized titanium abutment (control group, Straumann CARES abutments/Straumann BL implants [C]) were included. All abutments were one-piece abutments with an internal implant-abutment connection and were customized to the same shape but featured different implant-abutment connection designs. For each group, 20 identical copies of a master abutment were fabricated and fixed on their corresponding implants. Half of the abutments in each group were left unrestored, and the other 10 received glass-ceramic crowns. Static loading was applied at a 30-degree angle to the palatal surface until failure, and bending moments were calculated. The type of failure was characterized visually by dismounting the abutments and by examination of cross-sections of the embedded specimens. The results were analyzed statistically.
RESULTS: The mean range of bending moments was higher for the unrestored groups (158.2 to 678.2 Ncm) than for the restored groups (117.9 to 419.4 Ncm). The highest mean bending moments were seen in the control group, both restored and unrestored (419.4/678.2 Ncm). Unrestored, T1 and T2 exhibited significantly higher bending moments than T3. This was also observed in the restored groups. CONCLUSION:Both the abutment material and the implant-abutment connection design affected the bending moments of abutments after static loading. Internally connected zirconia abutments with horizontal mismatch to the implant exhibited significantly higher bending moments compared to those without horizontal mismatch.