This proposal is motivated by a need to understand the risk factors for tooth loss in middle age and older populations, and to supplement the efforts of NIDR in achieving the goal of """"""""eliminating the loss of teeth in U.S. populations."""""""" The causes of tooth loss are still unclear in older subjects. No quantitative estimates are available on the risk of loss for teeth which are differentially affected by various levels of dental caries, periodontal disease, or both at a given age. Moreover, data-driven definitions of the threshold of clinical variables such as number of carious or restored surfaces; extent of probing depth or bone loss at which the risk of loosing teeth increases significantly, are not available. In contrast, it is also important to characterize those teeth which are lost but no biological reasons ar.e apparent for their loss. The present proposal is uniquely capable to address these issues in a cost- effective manner. Fifteen years of longitudinal oral health data on 697 subjects available from the Boston VA Dental Longitudinal Study (DLS) will be analyzed. The oral health data include tooth and surface specific information on clinical parameters of dental caries, periodontal disease and prosthetic replacement. Radiographs were used to assess the proximal surfaces for caries and bone loss. Subject-based and tooth-based analysis will be carried out to determine the incidence of tooth loss and associated risk factors. Three longitudinal data analysis approach are proposed to analyze these data: l) Baseline variables predicting the future outcome, 2) Repeated measures design, using the linear growth curve model that falls in the framework of Laird and Ware random-effects longitudinal models, and 3) Pooling of repeated observations (PRO) model used in Framingham Heart Study. In addition, those teeth which are lost without any apparent biological reason will be characterized with respect to their relationship to other extracted teeth and new prosthesis. Analogous to many medical diseases the tooth-based logistic regression analysis would provide """"""""Life tables"""""""" for teeth differentially affected by various levels of caries and periodontal disease.
| Krall, E A; Dawson-Hughes, B; Garvey, A J et al. (1997) Smoking, smoking cessation, and tooth loss. J Dent Res 76:1653-9 |
