Multiple regression analysis predicts the dynamic of chondrocytes stimulated by magnetic and electric fields

Abstract

Purpose: The aim of this study was to implement a multiple regression analysis to find mathematical models that estimate the proliferative rate and the molecular synthesis of chondrocytes when these cells are stimulated either by magnetic or electric fields. Methods: Data derived from previous studies performed in our laboratory were used for statistical analyses, which consisted of applying magnetic fields (1 and 2 mT) and electric fields (4 and 8 mV/cm) to chondrocytes. Data from cell proliferation and glycosaminoglycan expression were used to adjust and to validate each mathematical model. Results: The root square model efficiently predicted the chondrocyte dynamics, evidencing determination coefficients of R2 = 92.04 for proliferation and R2 = 70.95 for glycosaminoglycans when magnetic fields were applied, and R2 = 88.19 for proliferation and R2 = 74.79 for glycosaminoglycans when electric fields were applied. Conclusions: The reduced, interactive, quadratic and combined models exhibited lower R2 values, nevertheless, they were useful to predict proliferation and glycosaminoglycan synthesis, as the right-skewed distribution, determined by the F parameter, evidenced a Frejected < Fcomputed. The models are efficient since the prediction of chondrocyte dynamics is comparable to the cell growth and to the molecular synthesis observed experimentally. This novel formulation may be dynamic because the variables that fit the models may be modified to improve in vitro procedures focused on cartilage recovery.