The study of pituitary-thyroid hormones ratio based on regression analysis

Abstract

Aim. To work up a mathematical model for calculation of the levels of pituitary-thyroid hormones system by least squares method in rats with experimental hypothyroidism.
 Methods. To study the relationship of hormones levels regression analysis was used. The search of coefficients was performed using the least squares method. Investigations were carried out on rats, which were divided into six groups of 12 rats each: the first group was control, in the animals of second, third, fourth, fifth and sixth groups hypothyroidism was induced by daily intragastric administration of tiamazol in following doses: 2.5; 20.0; 10.0; 5.0 and 1.0 mg per 100 g of rat body weight for 3 weeks.
 Results. Regression analysis was carried out, the type of regression as well as parameters were chosen; statistical analysis of the relationship of hormones was conducted based on the obtained results. By comparing the calculation results of the laboratory analysis sufficiently high reliability of the developed model was set up. The deviation of the arithmetic mean value of the level of thyroid stimulating hormone, produced on the basis of experimental and calculated data, is 2.7%. The accuracy of thyroid stimulating hormone levels calculations increased with the decrease of thyrostatic medication dose. The relative error while calculating the free thyroxine levels in the same groups of rats did not exceed 2.15%, accounting for 1.64; 1.34; 0.36 and 2.15%, respectively, when administered daily 20.0; 10.0; 5.0 and 1.0 mg of antithyroid drug per 100 g body weight. One can argue about the reliability of the constructed model to reproduce the performance levels of the hormones of the pituitary-thyroid system.
 Conclusion. In the absence of modern high sensitive immunochemiluminescent diagnostic methods the results can be used for thyroid-stimulating hormone levels calculation as one of the major markers of the thyroid gland functional state.