Abstract
Purpose: The purpose of this study was to examine the risk of early childhood caries (ECC) in children who had middle ear infections (MEI) or respiratory tract infections (RTI) during early childhood. Methods: Medicaid data from Michigan were analyzed for all continuously enrolled children born in 2001 for whom enrollment, medical, and dental claims were filed during 2001-2004. Proportional hazards survival models were used to assess the risk of ECC in children who had MEI or RTI during the first year of life. Results: Included in the study were 29,485 children (51% males and 49% females). By first year of life, 47% and 69% of children had a claim for MEI and RTI, respectively. Children with at least one claim for MEI or RTI were at 29% higher risk for developing ECC compared to those with no claims (P<.001). Hispanic children with 8 or more claims showed 91% greater risk for developing ECC than those with less than 8 claims (P=.01). Conclusions: The occurrence of middle ear infections or respiratory tract infections during the first year of life is associated with a significantly increased risk for developing early childhood caries during subsequent years. Race and ethnicity are possible predictors for ECC in the studied models. (Pedìatr Dent 2008;30:10510) Received December 12, 20Oo / Last Revision June 12, 2007 / Revision Accepted June 13, 2007 Table 4 also gives hazard ratios (HR) for each covariate in each model. A hazard ratio greater than 1 indicates that the variable is associated with an increased risk for ECC. A ratio less than 1 indicates a decreased risk for ECC, and a HR of 1 indicates no association. When each of the medical diagnoses (AOM, RTI, or UTI) was individually included in a proportional hazards model, AOM and RTI were individually significantly associated with the risk of ECC. The occurrence of AOM during the first year of life was associated with an 11% increase in the risk of ECC (HR=LIl, P=OS)1 and RTI was associated with a 34% increase (HR=1.34, -P<.001). Because the number of children with UTI was extremely small and UTI was not found to be significantly associated with ECC (HR=1.02, P=.9D, all models summarized in Table 4 were exclusively tested for the effects of AOM and RTI. This was done by including a variable called "AOM/RTI claims," which was coded as yes if the child had at least 1 claim for AOM or RTI during the first year of life, and no if the child did not have a diagnosis of AOM or RTI. In addition to the significant association between ECC and AOM/RTI claims as a dichotomous variable, Table 4 shows a dose-dependent response, where the strength of association was found to significantly increase as the frequency of AOM/RTI claims increased (Table 4, section C). The survival model in Table 4, section C was studied further by comparing the nature of the association between ECC and AOM/ RTI claims in children who had more frequent AOM/RTI claims to those with fewer or no AOM/RTI claims (section D). Children with frequent AOM/RTI claims were defined as those who had >8 claims for either AOM or RTI during the first year of life (group 4, section C). These children were compared to those who had <8 such claims (groups 1-3, section C). For all children combined (Caucasian, African American, and Hispanic), those who had >8 claims during the first year of life had a 30% higher risk for ECC compared to those with less frequent claims. When this model was stratified by race, a significant association between ECC and AOM/RTI claims was noted only in Hispanic children, who showed nearly a 91% increase in ECC risk for those who had more frequent AOM/ RTI claims.