Association Between Longitudinal Changes in Clinical and Radiographic Periodontitis Measures

Objectives: To examine the association between relative clinical attachment level (RCAL) loss and alveolar bone density (ABD) loss and alveolar bone height (ABH) loss over two years among postmenopausal osteopenic women.

Methods: Secondary analyses were performed on data from a two-year randomized, double-blind, placebo-controlled, clinical trial examining efficacy/safety of subantimicrobial dose doxycycline (SDD; 20 mg bid) in 128 postmenopausal osteopenic women on periodontal maintenance for moderate-advanced chronic periodontitis. RCAL loss was measured every 6 months by Florida probe (threshold for significant change: +/- 1.5 mm). Posterior vertical bite-wings were taken at baseline, one and two years for ABD (determined by computer-assisted densitometric image analysis [CADIA]) and ABH (Hausmann et al. method). Generalized Estimating Equations were used to fit regression models investigating the association between RCAL and alveolar bone measures.

Results: At baseline, RCAL loss was associated with ABH loss (p<0.0001). Periodontal disease status was relatively stable over 2 years with ABD gain/loss (6%/12% of sites) and ABH gain/loss (2%/8% of sites) occurring at a higher frequency than RCAL gain/loss (5%/3% of sites). One-year RCAL changes were not significantly associated with two-year changes in ABD (p=0.39) or ABH (p=0.87). The association between concurrent RCAL and ABD changes was modified by SDD treatment (p=0.037) where, among placebo subjects, the odds of ABD loss for sites with RCAL loss or no change were approximately twice the odds for sites with RCAL improvement (p=0.027); no significant association was seen among SDD subjects (p=0.66). ABH loss was not significantly associated with concurrent RCAL changes (p=0.58).

Conclusions: RCAL loss over two years does not appear to precede loss of ABH or ABD among postmenopausal osteopenic women. RCAL loss may occur concurrently with ABD loss, where SDD treatment may modify this association. Supported by NIDCR grants R01DE012872 and R03 DE019805-01A1.