All measurements were performed by one previously calibrated investigator.

The sample size calculation considered a standard normal distribution (type I error set at .05; type II error set at .20) and a sigma which was estimated based on the standard deviations observed in a recent preclinical animal study [4]. The clinical width of the alveolar ridge was defined as the primary outcome variable, considering a clinically relevant difference of 2 mm. A sample size of 15 patients per group was calculated to achieve a 95% power (Power and Precision, Biostat, Englewood, USA).

The statistical analysis of the pseudonymized data sets was accomplished using a commercially available software program (IBM SPSS Statistics 24.0, IBM Corp., Armonk, NY, USA).

Mean values, standard deviations, medians, 95% confidence intervals (CI), and frequency distributions were calculated for all outcomes assessed. The data rows were examined with the Shapiro-Wilk test for normal distribution. Between-group comparisons were accomplished using the unpaired t test. Linear regression analyses were used to depict the relationship between BI26 and CWb as well as SA26 values in both groups. The alpha error was set at 0.05.