The aim of this 10-year prospective study of 50-year-old individuals was to analyze the incidence of periodontal bone loss and potential risk factors for periodontal bone loss.Methods
The subject sample was generated from an epidemiological survey performed in 1988 of subjects living in the County of Värmland, Sweden. A randomized sample of 15% of the 50-year-old inhabitants in the county was drawn. At the 10-year follow-up in 1998, 320 (75%) of the 449 individuals examined at baseline were available for re-examination, out of which 4 had become edentulous. Full-mouth clinical and radiographic examinations and questionnaire surveys were performed in 1988 and 1998. Two hundred and ninety-five individuals (69%) had complete data for inclusion in the analysis of radiographic bone changes over 10 years. Non-parametric tests, correlations and stepwise multiple regression models were used for statistical analysis of the data.Results
The mean alveolar bone level (ABL) in 1988 was 2.2 mm (0.05) and a further 0.4 mm (0.57) (p = 0.000) was lost over the 10 years. Eight percent of the subject sample showed no loss, while 5% experienced a mean bone loss of ≥1 mm. Smoking was found to be the strongest individual risk predictor (RR = 3.2; 95% CI 2.03–5.15). When including as smokers only those individuals who had continued with the habit during the entire 10-year follow-up period, the relative risk was slightly increased (3.6; 95% CI 2.32–5.57). Subjects who had quit smoking before the baseline examination did not demonstrate a significantly increased risk for disease progression (RR = 1.3; 95% CI 0.57–2.96). Stepwise multiple regression analysis revealed that smoking, % approximal sites with probing pocket depth ≥4 mm, number of teeth and systemic disease were significant explanatory factors for 10-year ABL loss (R2 = 0.12). For never smokers, statistically significant predictors were number of teeth, mean ABL, % periodontally healthy approximal sites and educational level (R2 = 0.20).Conclusion
The inclusion of smokers in risk analysis for periodontal diseases may obstruct the possibility to detect other true risk factors and risk indicators.