Detection of metabolic syndrome with ATR-FTIR spectroscopy and chemometrics in blood plasma

• Published in: Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy Vol. 288; p. 122135
• Main Authors: Mateus Pereira de Souza, Nikolas, Hunter Machado, Brenda, Koche, Andreia, Beatriz Fernandes da Silva Furtado, Lucia, Becker, Débora, Antonio Corbellini, Valeriano, Rieger, Alexandre
• Format: Journal Article
• Language: English
• Published: England Elsevier B.V 05.03.2023
• Subjects: Ataxia Telangiectasia Mutated Proteins; ATR-FTIR spectroscopy; Blood plasma; Cell-Free Nucleic Acids; Chemometrics; Humans; Leptin; Metabolic syndrome; Metabolic Syndrome - diagnosis; Orthogonal partial least squares discriminant analysis; Plasma; Spectroscopy, Fourier Transform Infrared - methods; ATR-FTIR spectroscopy; Orthogonal partial least squares discriminant analysis; Metabolic syndrome; Chemometrics; Blood plasma
• DOI: 10.1016/j.saa.2022.122135
• ISSN: 1386-1425

[Display omitted] •Pioneering study to differentiate metabolic syndrome by ATR-FTIR in blood plasma.•p < 0.05 for control vs MetS (cm−1): 1717–1703, 1166–1137, 1113–1040, and 1027–1008.•100% accuracy with OPLS-DA model to discriminate patients with metabolic syndrome.•Amide I and amide II had the largest contributions to OPLS-DA loadings.•CRP and leptin (p < 0.05), and cfDNA (p < 0.01) for control vs metabolic syndrome. Metabolic Syndrome (MetS) is a constellation of 3 or more risk factor (abdominal obesity, high triglycerides, low HDL-c, high blood pressure, and elevated blood glucose) for atherosclerotic cardiovascular disease. Considering these systemic metabolic changes in the biochemical pathways of all biomolecules, Attenuated Total Reflection-Fourier Transform Infrared (ATR-FTIR) spectroscopy is a rapid, low-cost, and reagent-free alternative technique capable of identifying spectral biomarkers that differentiate subjects with MetS from control. In this study, plasma samples from 74 subjects (14 MetS, 60 control) were analyzed on the ATR-FTIR spectrophotometer. The objective was to differentiate subjects with MetS from control with supervised chemometrics modeling (Orthogonal Partial Least Squares-Discriminant Analysis, OPLS-DA). Additionally, the inflammatory status of subjects with MetS and control (supervised by C-reactive protein - CRP, leptin, and cell-free DNA - cfDNA) was verified. The OPLS-DA model achieved 100% sensitivity and specificity in cross-validation. For 1 latent variable (93.4% of variance), RMSECV < 0.002, PRESS CV < 0.0001, and R2 > 0.9999 was obtained. Significant spectrochemical differences (p < 0.05) were found between MetS and control subjects in the following biomolecular regions (cm−1): 1717–1703 [ν(CO) and δ(NH)], 1166–1137 [ν(C-OH) + ν(CO) and ν(CC) + δ(OH) + ν(CO)], 1113–1040 [ν(PO2-) and ν(C-OH)], and 1027–1008 [ν(CO) and v(CH2OH)]. In the OPLS-DA model loadings, amide I [1720–1600 cm−1, ν(CO)] and amide II [1570–1480 cm−1, δ(NH) + ν(CH)] had significantly greater weight than all other regions. There was a significant difference in inflammatory status between MetS patient and control (p < 0.05 for CRP and leptin, and p < 0.01 for cfDNA).