Periodontitis, a chronic inflammatory disease of the periodontium, affects an estimated 50% of US adults over the age of 35. To develop appropriate planning for treatment and cure, dental health care professionals must understand the biological, socio- demographic, behavioral and other medical factors that affect tooth loss through periodontal progression. However, the statistical methods employed to understand these come with several interesting challenges. The data are multivariate, non-Gaussian, non- stationary and have missing information which are informative of the oral health status of that oral region. Besides, one can conjecture that periodontal progression can be spatially referenced. Current statistical methods do not address all of these under a unified framework. Goals: Using a Bayesian paradigm, the proposed study will develop robust statistical methods combining all the above challenges, for assessing periodontal disease status and identifying important covariates that are associated. Subjects: The statistical methods will be evaluated on a dataset of 313 dentate subjects who were enrolled in the Gullah African-American (AA) Diabetics (GAAD) Study as part of the SC COBRE for Oral Health. For generalizability, the methods will be investigated on nationally-representative data collected as part of NHANES (1999-2004). Available data and study design: Periodontal status (determined by site-level pocket dept, clinical attachment level, and bleeding on probing), other relevant biological and medical status like smoking habits, brushing and flossing habits, demographics (poverty status) and other parameters have been collected at the Medical University of South Carolina (MUSC) as part of the GAAD study. The Gullah-AA subjects represent an interesting population with minimal genetic admixture whose dental health status remains vastly unknown. NHANES data are publicly available. Significance: The new statistical methods will provide dental researchers enhanced knowledge about how spatial associations might predict periodontal progression in presence of the aforementioned characteristics typical for periodontal data. This will enable researchers to better target risk assessment and prevention strategies, thereby improving health status.
Data collected during a periodontal examination are complex, with several types of responses taken at various mouth locations, complicated missing data patterns, and might have spatially referenced disease progression. Standard methods currently used do not consider these complications and might provide inefficient and/or biased information. This project proposes robust statistical models addressing these concerns, has the potential to exert tremendous impact from a dental public health perspective, and will bridge the gap between dental practitioners and academic dental basic science researchers for developing novel prevention strategies. The long term goal is to provide dental researchers a better understanding about repeated and longitudinal dental (health) data, so as to prevent and control disease and improve overall dental health.
| Bandyopadhyay, Dipankar; Galvis, Diana M; Lachos, Victor H (2017) Augmented mixed models for clustered proportion data. Stat Methods Med Res 26:880-897 |
| Lewis, Bradley R; Bandyopadhyay, Dipankar; DeSantis, Stacia M et al. (2017) Augmenting beta regression for periodontal proportion data via the SAS NLMIXED procedure. J Appl Probab Stat 12:49-66 |
| Cai, Bo; Bandyopadhyay, Dipankar (2017) Bayesian semiparametric variable selection with applications to periodontal data. Stat Med 36:2251-2264 |
| Jin, Ick Hoon; Yuan, Ying; Bandyopadhyay, Dipankar (2016) A BAYESIAN HIERARCHICAL SPATIAL MODEL FOR DENTAL CARIES ASSESSMENT USING NON-GAUSSIAN MARKOV RANDOM FIELDS. Ann Appl Stat 10:884-905 |
| Schnell, Patrick; Bandyopadhyay, Dipankar; Reich, Brian J et al. (2015) A marginal cure rate proportional hazards model for spatial survival data. J R Stat Soc Ser C Appl Stat 64:673-691 |
| Matos, Larissa A; Bandyopadhyay, Dipankar; Castro, Luis M et al. (2015) Influence assessment in censored mixed-effects models using the multivariate Student's-t distribution. J Multivar Anal 141:104-117 |
| Galvis, Diana M; Bandyopadhyay, Dipankar; Lachos, Victor H (2014) Augmented mixed beta regression models for periodontal proportion data. Stat Med 33:3759-71 |
| Mutsvari, Timothy; Bandyopadhyay, Dipankar; Declerck, Dominique et al. (2013) A multilevel model for spatially correlated binary data in the presence of misclassification: an application in oral health research. Stat Med 32:5241-59 |
| Bandyopadhyay, Dipankar (2013) From Mouth-level to Tooth-level DMFS: Conceptualizing a Theoretical Framework. J Dent Oral Craniofac Epidemiol 1:3-8 |
| Reich, Brian J; Smith, Luke B (2013) Bayesian quantile regression for censored data. Biometrics 69:651-60 |
Showing the most recent 10 out of 16 publications
