In vitro marginal leakage in varnished and lined amalgam restorations

Black's class I classic cavity preparations were completed in 440 extracted, intact human premolars and restored with two silver amalgams and five different base materials with and without cavity varnish, which resulted in 20 combinations. A standard fluorescent dye test was performed and the percentage of marginal leakage recorded after storage for 7 days, 3 months, and 1 year in 1% NaCl at 20 degrees C. The results were computed by use of a linear logistic statistical analysis with a critical level of statistical significance of p < 0.05. Sealing improved with time and there was consistently better sealing in the first months without varnish, but after 1 year this was not significant. The low-copper amalgam recorded better sealing than the high-copper amalgam. A calcium hydroxide base promoted sealing but base-free cavities performed second best. Sealing declined with 1 year if Dycal base was used. Zinc-containing bases exhibited poor sealing, even those containing tin. Base solubility can improve sealing in calcium-containing bases but not in zinc-containing bases.