A New Regression Model Based on Salivary Biomarkers, Chronological Age and Gender to Predict the Stage of Cervical Vertebral Maturation in Orthodontic Patients.

Abstract

Background: In orthodontics, radiography-based methods are frequently used for the assessment of skeletal maturity. Concerning X-ray exposure hazard, a new objective and less invasive method is needed to identify the optimal orthodontic treatment timing.

Purpose: In this study, the pattern of salivary alkaline phosphatase and insulin-like growth factor-1 in the circumpubertal period was investigated. Moreover, new regression models were built to predict skeletal maturity more accurately.

Materials and method: In this cross-sectional study, fifty-five subjects aged 7 to 20 years were enrolled to compare the levels of salivary alkaline phosphatase and insulin-like growth factor-1 at different skeletal maturity stages by using 6-stage cervical vertebral maturation method. Moreover, three new multinomial logistic regression models based on these biomarkers, as well as chronological age, and gender were built to predict the cervical stage. Statistical analysis was performed using SPSS (version 24) software. In addition to descriptive statistics, the Shapiro-Wilk test, the one-way analysis of variance test, the independent Samples T-test, the Pearson correlation coefficient, the Chi-square test, and the overall correct classification rate was performed.

Results: A significant difference was observed for salivary alkaline phosphatase and also insulin-like growth factor-1 levels between cervical stages (< 0.001). The peak level in both salivary alkaline phosphatase and insulin-like growth factor-1 levels was observed at CS3. The levels of these biomarkers had a significant positive correlation with the cervical stages from CS1 to CS3 and a significant negative correlation from CS3 to CS6. Based on the regression model analysis, we found that Model 3´ which combined gender with chronological age, alkaline phosphatase, and insulin-like growth factor-1 had the highest predictive ability (76.4%).

Conclusion: The notably larger predictive ability of the new model which combined gender with chronological age, salivary alkaline phosphatase, and salivary insulin-like growth factor-1 might be responsible for the identification of the optimal orthodontic treatment timing in an objective and less invasive manner in healthy growing individuals.