abstract
Given the recognized lack of prenatal clinical methods for the early diagnosis of preterm delivery, intrauterine growth restriction, preeclampsia and gestational diabetes mellitus, and the continuing need for optimized diagnosis methods for specific chromosomal disorders (e.g., trisomy 21) and fetal malformations, this work sought specific metabolic signatures of these conditions in second trimester maternal urine, using 'H Nuclear Magnetic Resonance (H-1 NMR) metabolomics. Several variable importance to the projection (VIP)- and b-coefficient-based variable selection methods were tested, both individually and through their intersection, and the resulting data sets were analyzed by partial least-squares discriminant analysis (PLS-DA) and submitted to Monte Carlo cross validation (MCCV) and permutation tests to evaluate model predictive power. The NMR data subsets produced significantly improved PLS-DA models for all conditions except for pre-premature rupture of membranes. Specific urinary metabolic signatures were unveiled for central nervous system malformations, trisomy 21, preterm delivery, gestational diabetes, intrauterine growth restriction and preeclampsia, and biochemical interpretations were proposed. This work demonstrated, for the first time, the value of maternal urine profiling as a complementary means of prenatal diagnostics and early prediction of several poor pregnancy outcomes.
keywords
ACID-CONCENTRATIONS; PRENATAL DISORDERS; VARIABLE SELECTION; METABOLIC PROFILE; OXIDATIVE STRESS; AMNIOTIC-FLUID; DOWN-SYNDROME; METABONOMICS; BIOMARKERS; PLASMA
subject category
Biochemistry & Molecular Biology
authors
Diaz, SO; Barros, AS; Goodfellow, BJ; Duarte, IF; Galhano, E; Pita, C; Almeida, MD; Carreira, IM; Gil, AM
our authors
Ana Maria Pissarra Coelho Gil
Associate Professor with Aggregation
António de Sousa Barros
Post-doc Fellowship
Brian James Goodfellow
Assistant Professor
Iola Melissa Fernandes Duarte
Principal ResearcherProjects
acknowledgements
The group acknowledges funding from the European Regional Development Fund-FEDER through the Competitive Factors Thematic Operational Programme-COMPETE and the Foundation for Science and Technology-FCT, Portugal (PEst-C/CTM/LA0011/2011, PEst-C/QUI/U10062/2011, PTDC/QUI/66523/2006). S.O.D. thanks FCT for SFRH/BD/64159/2009 grant. A.M.G. acknowledges the Portuguese National NMR Network (RNRMN), supported with FCT funds, and M. Spraul, Bruker BioSpin, Germany, for access to software and spectral databases.