Análise de plasma sanguíneo por espectroscopia ATR-FTIR: identificação de biomoléculas em mulheres com diabetes mellitus gestacional e suas implicações em desfechos adversos neonatais

Abstract

Gestational Diabetes Mellitus (GDM) is a hyperglycemic imbalance, which affects 5 - 25% of pregnancies worldwide. Currently, there is a lack of consensus on the knowledge of the pathophysiology of GDM. The aim of this study is to propose the use of Fourier Transform Infrared Spectroscopy with Total Attenuated Reflectance (ATR-FTIR), using blood plasma samples to characterize the GDM. The study was conducted with two groups: control (n = 50) and women diagnosed with GDM (n = 50), attended at two public maternity hospitals. The blood plasma was inserted in the ATR-FTIR for analysis of the fingerprint in the region of biological interest (1,800 - 900 cm-1 ), analyzed according to the unsupervised models PCA (Principal Component Analysis) and supervised SPA (Projection Algorithm Successive) and GA (Genetic Algorithm), together with LDA (Linear Discriminant Analysis), QDA (Quadratic Discriminant Analysis), SVM (Support Vector Machines) algorithms. The results obtained by GA-LDA were the most satisfactory, using only chosen spectral variables that could be related to chemical groups of different biomolecules. The GALDA models correctly classified the DMG group vs the Control group, with a sensitivity and specificity of 96.7% and attributed the main discriminators between the classes of lipid, carbohydrate and protein biomolecules (region from 1,500 to 1,800 cm-1 ). It is demonstrated the potential of ATR-FTIR techniques and multivariate models as possible methodologies for the characterization of GDM, we suggest that the increase of these peaks in the region of protein biomolecules may be related to the pathophysiology of GDM. It was observed that newborns who presented macrosomia were associated with women with GDM, who had lower protein density in the regions 1,587 cm-1 and 1,589 cm-1 . The combination of patient-specific information with blood biomarker data opens a new perspective to investigate the mechanisms in GDM