Color and translucency of aged and stained composite materials

Objectives: To evaluate color and translucency of composite resin restorative materials subjected to aging and staining.
Methods: Spectral reflectance (SR) and color values of Kerr Herculite Ultra (H) and GC Kalore (G) at various thicknesses were determined on white, gray and black backings using a spectroradiometer and an external light source. Specimens were subjected to aging in total darkness at room temperature for 2 years and colors were re-determined. Specimens were then themocycled between hot and cold coffee and colors were determined again. The error of Kubelka-Munk theory (KMT) was analyzed using 3-way ANOVA at selected wavelengths. CIEDE2000 color difference was calculated between original and aged specimens and between aged and stained specimens. Relative Translucency Parameter (RTP) for original, aging alone and aging and staining was determined as the color difference over white and black backings. CIEDE2000 and RTP values were analyzed by ANCOVA with thickness as the covariate.
Results: Analysis of error of KMT: The 2 way interactions of Brand*Time, Brand*Wavelength and Time*Wavelength were statistically significant (p≤0.042) with staining generally producing significantly higher error values.
Regarding color difference due to aging and staining: The 2 way interactions of Brand*Time and Shade*Time were statistically significant (p≤0.014). Brand G stained more than Brand H for each shade studied. For each Shade and Brand, the color change due to staining was significantly (p≤0.005) higher than the color change due to aging.
Regarding the Relative Translucency Parameter (RTP): The 2-way interactions between Brand*Condition and Brand*Shade were found to be statistically significant (p≤0.018). With the exception of Brand H Shade C2, RTP was significantly different due to aging or aging and staining.
Conclusions: Stained specimens should not be fit to KMT due to high error values. Optical properties of stained composite materials may be analyzed by ANCOVA with thickness as the covariate.