Crestal bone changes in tissue‐level implants according to implant placement depth: A retrospective cohort study

BACKGROUND: This study evaluated the effect of tissue-level implant placement depth on crestal bone changes before and after loading and explored patient-, implant-, and prosthesis-related factors associated with these changes. METHODS: This retrospective cohort study included patients with at least one tissue-level implant. Implants were classified by placement depth (crestal, partially, or completely subcrestal). Crestal bone levels were assessed radiographically at placement, pre-loading, and post-loading, and categorized relative to the smooth collar or implant body. Clinical records provided patient, implant, and prosthesis data. Associations were analyzed using mixed-effects logistic regression model. RESULTS: A total of 183 implants (20 crestal, 100 partially subcrestal, and 63 completely subcrestal) were analyzed in 96 patients. Crestal implants showed no pre-loading crestal bone change and remained stable post-loading. In subcrestal implants, bone change was mainly confined to the smooth collar. Completely subcrestal implants showed higher odds of exhibiting a more apically located pre-loading crestal bone change (adjusted odds ratio [OR] = 2.44, p = 0.045) than partially subcrestal implants. Among implants without pre-loading bone change, completely subcrestal placement was associated with a higher probability of post-loading bone change compared with partially subcrestal placement (91.8% vs. 49.2%; adjusted OR = 11.56, p = 0.035). No other variables were significantly associated with these outcomes (p > 0.05). CONCLUSIONS: Placement depth was a key factor associated with crestal bone changes in tissue-level implants. Subcrestal placement was associated with greater pre-loading crestal bone change in a depth-dependent manner, with the highest risk of post-loading change observed in completely subcrestal implants without prior bone remodeling. Both pre- and post-loading changes were predominantly confined to the smooth collar.