IADR Abstract Archives

Finishing Effects on Cement Surfaces at Different Marginal Fit Levels

Subtractive techniques to refine the form of exposed marginal cement surface require a minimum amount of excess. Presence of cement gaps at restoration margins create shelter from shear force, where plaque can accumulate.


Purpose of this in vitro study was to clarify how two bi-dimensional parameters of marginal cement deficiency (CD) are affected by the level of marginal fit (MF), the selected luting agent (LA) and finishing technique (FT).


With standardized procedures each of 120 identical two-component samples was prepared with a combination of the two-level independent variables MF, LA and FT forming 8 groups. CD area (CDA) in μm2 and CD solidity (CDS), a shape descriptor, were evaluated in a randomized, 2x2x2 factorial and, balanced experimental design (n=15). The negative replicas, produced from cemented samples, were stained and intersected along their long axis into six pieces. Digital images of the sections captured in a stereoscopic microscope, were evaluated with an image analysis program. Data from each response variable were analyzed with 3-way ANOVA and Tamhane T2 post hoc tests (α=.05).


The next table summarizes F-ratios with associated degrees of freedom and significant values, for main and interaction effects on each response variable:

Main and interaction effects

CDA (in μm2)



F(1,88)=208.428, P<.0001

F(1,96)=0.140,   P=.709 (ns)


F(1,88)=0.563,    P=.455 (ns)

F(1,96)=14.856,  P<.001


F(1,88)=62.098,   P<.0001

F(1,96)=30.712,  P<.0001


F(1,88)=0.672,   P=.415 (ns)

F(1,96)=3.021,   P=.085 (ns)


F(1,88)=66.493,  P<.0001

F(1,96)=14.128,  P<.001


F(1,88)=15.961,  P<.001

F(1,96)=21.073,  P<.0001


F(1,88)=16.368,  P<.001

F(1,96)=0.919,   P=.340 (ns)

ns=not significant


Within the limitations of this study, the following can be concluded: Cement deficiency solidity is affected by the selected luting agent. Initial marginal discrepancy affects the cement deficiency area and the finishing technique affects both response variables. Additionally, multiple interaction effects between the independent variables were efficiently revealed with the factorial design.

Pan European Region Meeting
2014 Pan European Region Meeting (Dubrovnik, Croatia)
Dubrovnik, Croatia
Scientific Groups
  • Villias, Aristeidis  ( University of Cologne, Cologne, , Germany )
  • Niedermeier, Wilhelm  ( University of Cologne, Cologne, , Germany )
  • Poster Session
    Prosthodontics III