Prediction of Periodontitis Occurrence: Influence of Classification and Sociodemographic and General Health Information |
| |
| Authors: | Fábio R.M. Leite Karen G. Peres Loc G. Do Flávio F. Demarco Marco A.A. Peres |
| |
| Affiliation: | 1. Currently, Department of Dentistry and Oral Health, Section of Periodontology, Faculty of Health Sciences, Aarhus University, Aarhus, Denmark;2. previously, Postgraduate Program in Dentistry, Federal University of Pelotas, Pelotas, Rio Grande do Sul, Brazil.;3. Australian Research Center for Population Oral Health, School of Dentistry, The University of Adelaide, Adelaide, SA, Australia.;4. Postgraduate Program in Epidemiology, Federal University of Pelotas.;5. Postgraduate Program in Dentistry, Federal University of Pelotas. |
| |
| Abstract: | Background: Prediction of periodontitis development is challenging. Use of oral health–related data alone, especially in a young population, might underestimate disease risk. This study investigates accuracy of oral, systemic, and socioeconomic data on estimating periodontitis development in a population‐based prospective cohort. Methods: General health history and sociodemographic information were collected throughout the life‐course of individuals. Oral examinations were performed at ages 24 and 31 years in the Pelotas 1982 birth cohort. Periodontitis at age 31 years according to six classifications was used as the gold standard to compute area under the receiver operating characteristic curve (AUC). Multivariable binomial regression models were used to evaluate the effects of oral health, general health, and socioeconomic characteristics on accuracy of periodontitis development prediction. Results: Complete data for 471 participants were used. Periodontitis classifications with lower thresholds yielded superior predictive power. Calculus, pocket, or bleeding presence at age 24 years separately presented fair accuracy. Accuracy increased using multivariable models; for example, the Beck et al. classification AUC from 0.59 to 0.75 combining proportion of teeth with calculus, bleeding, or pocket with income; number of lost teeth; sex; education; people living in the house; prosthetic needs; or number of decayed, missing, or filled teeth (DMFT). Proportion of teeth with pocket, bleeding, or calculus; number of DMFT; toothbrushing frequency; blood pressure; sex; and income were most frequently associated. Conclusions: Choice of classification might have an impact on accuracy to predict periodontitis occurrence. Regardless of the classification, predictive value for development of periodontitis in young adults might be increased by combining periodontal information, sociodemographic information, and general health history. |
| |
| Keywords: | Cohort studies epidemiology forecasting incidence periodontal diseases periodontitis |
|
|
