Over the 5-year study period, we report < 2.0 mm bone level change for all implants, 0.1–0.5 mm for 40%, and no bone loss or bone gain for 38% of all implants. Additionally, bone loss was 0.32 ± 0.66 mm and 0.13 ± 0.29 mm for the platform-switching and platform-matching subgroups. Of note, in this study, the platform-matching and platform-switching groups were very unbalanced (67 vs. 206 implants) because the decision to choose abutment type was the clinician’s choice according to the clinical situation.

Furthermore, very few radiographs were available for the platform-matching subgroup; thus, differences between the two subgroups are not conclusive. Nevertheless, the minimal crestal bone loss of 0.32 mm observed for platform-switching implants is comparable with the data reported in other studies on platform-switching implants [17, 23]. The bone gain of 0.12 ± 0.42 mm at 1-year follow-up [17] and of 0.16 ± 0.53 mm at 3 years follow-up [23] have been reported. In these studies, the outer geometry of the implant was comparable; however, Rocha et al. [23] used implants of the same kind while Moergel et al. [17] used implants with a conical connection.

The importance of the vertical soft tissue thickness has recently been reported [33, 34]. Platform-switching implants placed in thick tissues led to the preservation of the crestal bone level, while this was not observed in thin mucosal tissues. These studies were not yet published in the planning phase and initiation of the present study. Accordingly, pocket probing depth measurements were performed rather than vertical soft tissue thickness. These measurements may be biased; it is thought that the probe may stop at the horizontal shift instead of the pocket depth, yet, to our knowledge, there is no reference supporting this. In daily practice, probing was sometimes not performed if the implants showed no pathological findings.