Cuspal deflection can cause micro-crack propagation, crazing, reduction of fracture resistance, pulpal conditions, or even cusp fracture [8,27]. The present results revealed that the cuspal deflection of restored cavities with bulk-fill composites was significantly lower than that of cavities restored with the conventional composite with the incremental application technique at all-time points. Thus, the null hypothesis of the study in this respect was rejected. This finding was in agreement with the results of previous studies on this topic [28,30–32]. Different bulk-fill composites were not significantly different regarding cuspal flexure in our study, which was in accordance with previous findings. [27,31] A systematic review  indicated that despite the differences in the methodology of studies (type of composite, sample size, technique of testing), bulk-fill composites show lower cuspal deflection than conventional composite resins. Lower cuspal deflection of bulk-fill composites can be attributed to the structure of their resin matrix, and filler technology . Lower cuspal deflection of X-tra fil may be due to its higher filler content compared with P60 . TEGDMA-rich matrix of conventional composites results in higher cross-linking and higher polymerization shrinkage ; while, bulk-fill composites have higher amounts of UDMA and bis-EMA oligomers and no or very small amount of TEGDMA, resulting in lower polymerization shrinkage and lower cuspal deflection [31,32]. Higher UDMA content of bulk-fill composites, which is a high molecular weight monomer with relatively low viscosity, prolongs the polymerization reactions and leads to higher stress release due to higher composite flow before reaching the gel point [28,32]. The manufacturer of Xtra-fill composite claims that stress relievers have been used in its composition, serving as chemical cushions between the filler particles, and resulting in higher elasticity and lower polymerization shrinkage and cuspal deflection .
In the present study, maximum cuspal deflection was immediately after restoration, which significantly decreased within 7 days. These findings confirmed the previous results in this respect . Following water storage, internal stresses are gradually released due to hygroscopic expansion of composite . However, the primary stress generated in the first couple of minutes after restoration causes cuspal deflection, and may lead to debonding, enamel cracks, postoperative tooth hypersensitivity, and marginal microleakage. These side effects are not reversible even after compensation of polymerization shrinkage by hygroscopic expansion of composite.
Cuspal deflection of conventional composite restorations has been reported in laboratory experiments from 15 to 45 µm . Considering the limited number of previous clinical studies, a standard value for cuspal deflection that causes clinical problems or the effect of time of occurrence of these changes on development of clinical symptoms such as postoperative tooth hypersensitivity, pain or secondary caries have yet to be identified.
It has been reported that return of cuspal deflection to baseline value takes time, and may never occur in the medium-size and large cavities . In the present study, the values closely approximated the baseline value after 7 days, which was in agreement with the findings of a previous study . Width and depth of the cavity can affect cuspal deflection as well . Thus, similar mesio-occluso-distal cavities with the same dimensions were prepared in all teeth to eliminate the effect of this confounder on the results. Also, the type of light curing unit, curing time, light intensity, type of bonding system, and the bonding protocol were the same in all teeth to eliminate the effect of these confounding factors on the results.
In this study after 2000 thermal cycles marginal microleakage of restorations were assessed by the dye penetration technique using methylene blue because a significant correlation has been reported between the results of scanning electron microscopy and dye penetration test after 30 min of immersion in methylene blue . The results showed that microleakage was minimum at the occlusal enamel margin, followed by the proximal box enamel margin, which was in agreement with previous findings [37,41,42]. Histological, morphological and compositional differences between the enamel and dentin lead to different marginal adaptation and bond strength values .
One advantage of this study, compared with previous ones, was assessment of microleakage at both the occlusal and proximal box enamel margins, revealing that the occlusal enamel margin of restoration had significantly lower microleakage than the proximal box enamel margin. This finding can be due to higher degree of conversion of composite and adhesive at the occlusal margin due to shorter distance between the tip of the curing unit and the surface, compared with the proximal box enamel margin . Moreover, enamel thickness in the occlusal area is higher than that in the proximal area. Higher number of enamel cracks in the cervical region can also contribute to higher marginal leakage in the proximal area .
Microleakage is influenced by the type and size of tooth, C-factor of the cavity, adhesive technique, and composite polymerization technique . Since, we tried to standardize all these factors in the present study; therefore, the results obtained only reflect the effect of the composite type on microleakage. We found that irrespective of the location of margin, the difference in marginal microleakage was not significant between bulk-fill composites and the conventional composite applied incrementally, which was in line with previous findings [45–48] although the frequency of margins with no or low-grade leakage was higher in bulk-fill composites. Previous studies did not report a significant difference in marginal leakage of bulk-fill and conventional composite resins either [46–49]. Also, this study showed lower microleakage at the enamel compared with dentin margin, which was in agreement with previous findings , and is due to simpler process of bonding to enamel, and the challenges encountered in bonding to dentin due to the differences in the degree of mineralization and water content of enamel and dentin. In this study, maximum leakage was recorded for P60 conventional composite applied as bulk-fill in 4-mm increments. Marginal microleakage of other composite groups was not significantly different at any margin. Low marginal microleakage of bulk-fill composites can be due to their modified composition, higher curing depth, and lower polymerization shrinkage. Higher polymerization depth of bulk-fill composites is due to the type, size, and shape of filler particles and their chemical composition, resulting in their higher translucency and subsequently higher polymerization depth [28,35,43,56]. Addition of pre-polymerized fillers  and presence of stress inhibitors  in the composition of bulk-fill composites further decrease the polymerization stress. Also, some of bulk-fill composites have novel photo-initiators such as Ivocerin with higher capability to generate free radicals compared with camphorquinone .
This study was conducted on extracted teeth then lack of pulpal blood supply changes the physical and structural properties of dentin, which can affect the cuspal flexure and microleakage of composite resins after their polymerization. Short-term follow-up was another limitation of this study since cuspal flexure and microleakage were only assessed for 7 days after restoration. Future in-vivo studies are required to assess the effect of cuspal flexure following different restorative techniques particularly with bulk-fill composites on clinical signs and symptoms to obtain more reliable results.