A Linear Regression Prediction Model Of Early Childhood Caries Based On Caries Relevant Factors

ObjectiveEarly childhood caries (ECC) is that children younger than6-year-old have caries and the dmfs≧1. Because of its high incidence rate and severe damage, ECC has adverse effects on children oral local health and physical general development.The study of ECC’s etiology and pathogenesis can provide theoretical basis for prevention and treatment of ECC. It had been widely accepted that the dental plaque biofilm was the main etiological factor of caries. The analysis on the composition of microorganism in plaque from non-cariogenic to cariogenic was still limited. This study aimed to detect quantitatively the levels of4main cariogenic bacteria (S.mutans, S.sobrinus, L. acidophilus, A. naeslundii) and total bacteria in dental plaque and saliva samples by qRT-PCR, and to analyze the dmfs’association with the proportions of these cariogenic bacteria and the distribution regularity of the same cariogenic bacteria in plaque and saliva.Secretory immunoglobulin A (SIgA) was the main immunoglobulin in saliva and a major anticarious factor. SIgA could intervene and prevent the adhesion of bacteria to teeth by combining with the surface structure or surface antigen of bacteria, and then agglutinate and remove them. This study was designed to detect the SIgA level in saliva by ELISA, and to assess its relationship with dmfs and main cariogenic bacteria.In order to assess ECC risk more sensitively and effectively, in this study we tried to establish a linear regression prediction model of ECC based on cariogenic bacteria in plaque and saliva and salivary SIgA level.Methods Research objects are3-year-old children from3groups and each group has15children:caries-free group(dmfs=0), ECC group (dmfs=1～3) and S-ECC group (dmfs≧4).1)Dental plaques and saliva samples were obtained.2)Quantitatively detect4major cariogenic bacteria in dental plaque and saliva samples by real-time PCR based on SYBR Green I fluorescence and analyze pathogen proportions’association with dmfs and the percentage differences in plaque and saliva.3)Detect SIgA level in saliva by ELISA and assess its relationship with dmfs and the main cariogenic bacteria in saliva.4)Establish a prediction model of ECC based on the factors relevant to dmfs according to the research above.Results1. The relationship between the dmfs and cariogenic bacteria proportions in plaque and salivaThe proportions of S. mutans and L. acidophilus in dental plaque and the proportions of S. Sobrinus in dental plaque and saliva were significant positive correlated to the dmfs(P<0.01). The proportions of S. mutans in saliva and A. naeslundii in dental plaque were significant positive correlated to the dmfs (P<0.05).2. The relationship of the proportions of the same pathogen in plaque and salivaThe proportions of S. mutans, S. sobrinus and A. naeslundii in plaque and saliva were strongly positive correlated (P<0.05).3.The relationship between SIgA and dmfsSIgA showed significantly negative correlation to dmfs (P<0.01). There were significant differences of SIgA level among the three groups:CF>ECC (a<0.05), CF>S-ECC (a<0.05), but no significant difference between ECC and S-ECC.4.The relationship between SIgA and cariogenic bacteria in saliva samplesSIgA showed significantly negative correlation to S. mutans and S. sobrinus in saliva (P<0.01).5. A linear regression prediction model of ECC 1)4cariogenic bacteria in the plaque, S. mutans and S. sobrinus in saliva and salivary SIgA were significant predict factors to dmfs.2) Equation of linear regression: dmfs=69.483s.sm+49.977p.sm+48.845s.ss+63.634p.ss+175.076p.la+52.730p.an-0.087SIgA-11.2863)The adjusted variance R was0.948, meaning that the accuracy of the prediction model was94.8%.4)The F factor was strongly significant(P<0.001), showing that the prediction model had a extraordinary high specificity.Conclusions1.The proportions of S. mutans in dental plaque and saliva could be taken as a ECC early-warning index.2.S. sobrinus’percentage in dental plaque could assess the activeness of ECC and the proportions of S. sobrinus in dental plaque and saliva were highly positive corelated.3.L. acidophilus’percentage in dental plaque could assess the activeness of ECC. However, L. acidophilus in saliva might be a transient flora and undulate.4.A. naeslundii in dental plague might promote ECC.5.SIgA could inhibit the mutans streptococcus and prevent ECC, but as ECC aggravated, its anti-caries function became weaken.6.The linear regression prediction model established in this study displayed a specific and sensitive property and made it possible to screen ECC in the early stage.