Zero-inflated (ZI) models, which may be derived as a mixture involving a degenerate distribution at value zero and a distribution such as negative binomial (ZINB), have proved useful in dental and other areas of research by accommodating ‘extra’ zeroes in the data. Used in conjunction with generalised linear models, they allow covariate-adjusted inference of an exposure effect on the mixing probability and on the mean for the non-degenerate distribution. However, these models do not directly provide covariate-adjusted inference for the overall exposure effect. Focusing on the ZINB and ZI beta binomial models, we propose an approach that uses model-predicted values for each person under each exposure state. This ‘average predicted value’ method allows covariate-adjusted estimation of flexible functions of exposure group means such as the difference or ratio. A second approach considers a log link for both components of the ZINB to allow a direct approach to estimation. We apply these new methods to a study of dental caries in very low birth weight adolescents. Simulation studies show good bias and robustness properties for both approaches under various scenarios. Robustness diminishes when there is exposure group imbalance for a covariate with a large effect.

Objectives: The purpose of this study was to examine developmental enamel defects and dental caries in very low birth weight adolescents with high risk (HR-VLBW) and low risk (LR-VLBW) compared to full-term (term) adolescents. Methods: The sample consisted of 224 subjects (80 HR-VLBW, 59 LR-VLBW, 85 term adolescents) recruited from an ongoing longitudinal study. Sociodemographic and medical information was available from birth. Dental examination of the adolescent at the 14-year visit included: enamel defects (opacity and hypoplasia); decayed, missing, filled teeth of incisors and molars (DMFT-IM) and of overall permanent teeth (DMFT); Simplified Oral Hygiene Index for debris/calculus on teeth, and sealant presence. A caregiver questionnaire completed simultaneously assessed dental behavior, access, insurance status and prevention factors. Hierarchical analysis utilized the zero-inflated negative binomial model and zero-inflated Poisson model. Results: The zero-inflated negative binomial model controlling for sociodemographic variables indicated that the LR-VLBW group had an estimated 75% increase (p < 0.05) in number of demarcated opacities in the incisors and first molar teeth compared to the term group. Hierarchical modeling indicated that demarcated opacities were a significant predictor of DMFT-IM after control for relevant covariates. The term adolescents had significantly increased DMFT-IM and DMFT scores compared to the LR-VLBW adolescents. Conclusion: LR-VLBW was a significant risk factor for increased enamel defects in the permanent incisors and first molars. Term children had increased caries compared to the LR-VLBW group. The effect of birth group and enamel defects on caries has to be investigated longitudinally from birth.

Birth cohort studies of developmental defects of enamel (DDE) and early childhood caries (ECC) in very low birthweight (VLBW) and normal birthweight (NBW) infants are rare. In this birth cohort of 234 VLBW and 234 NBW infants, we report the incidence of ECC and DDE at 8 and 18-20 mos of corrected age. Infant medical and maternal socio-demographic data were abstracted from medical records at birth. Dental assessments for ECC and DDE (enamel hypoplasia, demarcated and diffuse opacities) were completed at 8 and 18-20 mos. The incidence of hypoplasia was significantly higher in VLBW compared with NBW infants (8 mos, 19% vs. 2%; 18 mos, 31% vs. 8%). The incidence of ECC (International Caries Detection and Assessment System: ICDAS ≥ 2) was 1.4% (8 mos) and 12% (18-20 mos) and was similar between the VLBW and NBW groups. At both ages, using a beta-binomial regression model to control for potential confounders (maternal and infant characteristics), we found increased risk for enamel hypoplasia among the VLBW infants compared with the NBW infants. African Americans had a lower risk for enamel hypoplasia at 18-20 mos. The VLBW infants should be monitored for ECC due to the presence of enamel hypoplasia.