Preoperative Salivary Active MMP-8 and Early Dental Implant Stability Trajectories: A Prospective Cohort Study

Objectives: We aim to investigate whether preoperative salivary active matrix metalloproteinase-8 (aMMP-8) is associated with the early dental implant-stability trajectory during the first 12 weeks after implant placement. Materials and Methods: This prospective, single-center cohort study included 125 adults receiving one implant each in healed edentulous sites. Unstimulated whole saliva was collected preoperatively for aMMP-8 assessment. Implant stability was assessed by resonance frequency analysis and expressed as mean implant stability quotient (ISQ) at placement (T0) and 4, 8, and 12 weeks. Longitudinal ISQ was analyzed using an unstructured-covariance mixed model for repeated measures with categorical time, adjusted for age, sex, smoking, diabetes status, periodontal status, salivary flow, jaw, implant dimensions, insertion torque, and bone quality. The primary analysis evaluated the global 3-df aMMP-8 × time interaction, followed by Holm-adjusted week-specific contrasts. Results: Of 500 scheduled ISQ measurements, 471 (94.2%) were observed. Six of 125 implants (4.8%) failed during the 12-week follow-up. Mean ISQ decreased from 75.9 ± 8.3 at T0 to 69.8 ± 7.5 at week 4 and subsequently increased to 71.4 ± 6.7 and 73.8 ± 7.6 at weeks 8 and 12, respectively. There was evidence of an aMMP-8 × time interaction (F(3, 112.1) = 8.81; p < 0.001). Each 1-SD higher preoperative ln(aMMP-8), equivalent to approximately a 1.89-fold higher concentration, was not clearly associated with ISQ change from T0 at week 4 (−0.81 ISQ units; 95% CI, −1.85 to 0.23; Holm-adjusted p = 0.126), but was associated with less favorable ISQ change at week 8 (−2.55; 95% CI, −3.75 to −1.35; Holm-adjusted p < 0.001) and week 12 (−3.09; 95% CI, −4.39 to −1.79; Holm-adjusted p < 0.001). The adjusted association at T0 was close to null (0.30; 95% CI, −1.07 to 1.66; p = 0.669), and the week-8 and week-12 findings were generally consistent across conventional sensitivity analyses, although the week-12 association was attenuated and no longer statistically significant under some failure-inclusive assumptions. Conclusions: Higher preoperative salivary aMMP-8 was associated with a less favorable early implant-stability trajectory, with differences emerging at 8 weeks and persisting at 12 weeks in the primary analysis. These findings support further evaluation of salivary aMMP-8 as a non-invasive preoperative marker of variability in early implant healing, although the clinical significance of the observed ISQ differences and their implications for implant outcomes remain uncertain.