Answers are grounded in this paper and cite their sources.
Marginal adaptation and fracture load are commonly evaluated as separate quality outcomes of adhesively bonded CAD/CAM restorations. A strong pooled correlation between them, however, may arise because material–cementation systems differ in both outcomes rather than because marginal gap independently tracks structural capacity within a fixed system. This secondary specimen-level analysis tested that distinction in a balanced factorial overlay experiment. Forty-eight maxillary first-premolar overlay specimens from a balanced 2 × 3 factorial experiment were analyzed. Restorations were fabricated from lithium disilicate or a resin-composite CAD/CAM block and cemented with adhesive resin cement, preheated restorative composite, or a sonically activated composite (n = 8/system). Mean marginal gap was measured before and after cementation, and fracture load was recorded by axial load-to-failure testing. Pearson/Spearman correlations, multivariable linear regression, six-system fixed effects, within-system centering, interaction testing, and HC3 robust inference were used. Post-cementation gap showed a strong crude inverse correlation with fracture load (r = −0.676; 95% CI − 0.806 to −0.486; p < 0.001). The crude slope was −85.0 N/µm (95% CI − 112.5 to −57.5), but after adjustment for material and cementation protocol it attenuated to −25.8 N/µm (95% CI − 69.8 to 18.2; p = 0.244). Material and cementation explained 58.0% of fracture-load variance; adding post-cementation gap increased R2 by 0.013. The within-system centered correlation was weak (r = −0.171; p = 0.244), whereas the descriptive correlation across six system means was strong (r = −0.960; p = 0.0024). The pooled inverse gap–load association attenuated markedly after adjustment for material and cementation. The exploratory within-system estimates were imprecise and do not exclude meaningful associations. Under the present in vitro conditions, marginal gap alone should not be regarded as an established predictor of fracture load.
No takes yet. Share an insight, caveat, or question.
Zaidoon Hasan Mohammed (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: