CBL mean values recorded around submerged and nonsubmerged implants at different timepoints are the most interesting results of the present randomized clinical trial. Before functional loading, radiographic CBL was significantly greater in submerged implants than that in nonsubmerged implants (0.23 mm ± 0.05 mm vs. 0.09 mm ± 0.07 mm). During the follow-up period, both implants showed similar changes in CBL that were not statistically significant (Δ = 0.12 ± 0.06 mm for submerged implants and 0.15 ± 0.09 mm for nonsubmerged implants). These results are in agreement with previously published comparative data between submerged and nonsubmerged implants on the short-term follow-up. A randomized clinical study by Juan Flores-Guillen et al. [14] reported at the end of the five-year follow-up a mean radiographic CBL of 0.59 (SD 0.92) mm and 0.78 (SD 1.03) mm for the submerged and nonsubmerged implants, respectively. Although differences of CBL at the end of the follow-up were not statistically significant, the submerged group demonstrated more crestal bone-level changes (− 0.52 mm versus − 0.24 mm) before functional loading (baseline to 6 months). Moreover, the authors indicated that more than half of the mean CBL reported at 5 years occurred during the healing of the implants and during the establishment of the biological width around the neck of the implant. In that study, however, for calculating the bone-level changes, the date of implant placement was used as a baseline, instead of the placement of the final restoration. Comparative data of peri-implant crestal bone changes between submerged and nonsubmerged implants, using as baseline the placement of the final restoration, were instead reported by Cecchinato et al. [6, 15]. At the end of the 5-year follow-up, the difference in CBL around submerged and nonsubmerged implants was not significant. However, between the baseline and the first year examination, the mean marginal bone loss for submerged implants was 0.17 (SD 0.5) mm versus 0.02 (SD 0.38) for nonsubmerged. Similar results related to CBL around submerged and nonsubmerged implants have been reported also by Siadat et al. [16] and Gheisari et al. [17]. In both studies, before loading, compared with submerged implants, nonsubmerged implants showed less CBL, but after 6 and 12 months of function, no significant differences were noted. A possible explanation of the greater CBL before loading around submerged implants could be related to the histological process of bone repair after the detachment of the periosteum carried out in the second surgical procedure, that in the one-stage surgical protocol is avoided [3]. Another possible explanation is the presence of an interface/micrograp at or below the alveolar crest in the submerged group. Histological analyses in animals have documented that bone-level implants placed in submerged and non-submerged (connected to healing abutments) approaches have a similar amount of bone loss [18]. A physiologic reaction to the presence of a microgap/interface seems to be connected to the microbial contamination at the microgap/interface [19,20,21], which in turn is associated with a significant inflammatory cell infiltrate [22]. In comparison, the complete absence of such microgap/interface produces no inflammation and consequently, no bone loss. Moreover, the magnitude of inflammation is proportionally dependent on the microgap/interface position relative to the alveolar crest. Subcrestal or crestal implant abutment microgap/interface promoted a significantly greater density of inflammatory reaction correlated with bone loss compared to supracrestal interfaces [22].