BACKGROUND: Limited subject-based information exists on the relationship between clinical and radiographic periodontal data.

AIMS: To use subject-based data to assess the extent of concurrence between clinical and radiographic information, and to study what clinical information best predicts alveolar bone loss (ABL). Material and Methods: Subject-based data on smoking habits, bleeding on probing, plaque scores, pocket probing depth (PD), and evidence of alveolar BL were obtained, and functional periodontal pentagon risk diagrams (PPRDs) were studied from 168 consecutive subjects attending a medical clinic.

RESULTS: The mean age of the subjects was 62.7 years (SD+/-9.0). The average number of teeth was 21.3 (SD+/-8.0) with on average 5.6 molars remaining (SD+/- 3.9). In this subject cohort, 33.1% had never smoked, 44.2% had quit smoking, and 22.7% were currently smokers. Mean plaque and bleeding scores were high or 60.2% (SD+/-24.0) and 53.1% (SD+/-23.6), respectively. PDs >or=6.0 mm were found in 55.9% of the subjects. Binary logistic regression analysis demonstrated that tooth loss and proportional plaque scores were the predominant factors included in the equations associated with ABL. Wald coefficients varied between 3.99 and 9.15, and with p-values between 0.05 and 0.01. When included, the PPRD score became the exclusive factor at several cut-off levels (Wald's coefficients between 19.8 and 15.6, p<0.001). Consequently, the best receiver operator curve was identified for the PPRD at the >40% cut-off ABL level (area under the curve: 0.81; 95% CI: 0.74-0.89; p<0.001).

CONCLUSIONS: The number of teeth lost and the proportion of plaque scores provided significant predictive factors for ABL. The functional PPRD demonstrated an exclusive and highly predictable association with ABL. Subject-based proportional data for PDs >4.0 mm provided poor substitute measures for the extent of ABL.