Exploring metabolomics through NMR: from early life to aging

Metabolomics occupies a unique yet demanding position in translational research. Its proximity to phenotype and its sensitivity to clinical and environmental perturbations make it a powerful tool for understanding biological processes in real-world settings. At the same time, this very sensitivity increases its vulnerability to confounding, analytical variability, and overinterpretation. This thesis addresses this dual nature by exploring how metabolomic evidence can be made more robust and credible, not through isolated improvements, but by aligning study design, analytical workflow, statistical inference, and reporting within a single, coherent framework. Early life is adopted as a particularly stringent translational model. Neonatal metabolism is highly dynamic, biological heterogeneity is substantial, and non-invasive biofluids such as urine provide a complex yet informative window into both systemic physiology and developmental processes, including renal maturation. These features make early life an ideal, though challenging, context in which to test the reliability and interpretability of metabolomic approaches. The thesis is organized into six chapters, following a progression from conceptual foundations to applied studies and broader reflections. Across these chapters, the thesis uses distinct biological contexts, spanning early life and aging, to examine how NMR-based metabolomics can support translational interpretation across different phases of the life course. Chapter 1 introduces metabolomics in the context of translational research, with a specific focus on early-life physiology, neonatal biofluids, longitudinal design, and multi-omics integration. Chapter 2 provides the methodological background necessary to understand the analytical approaches used in the thesis. The methodological framework is then applied to two neonatal studies. Chapter 3 presents an exploratory probiotic supplementation study in healthy, term, formula-fed infants. By integrating fecal microbiome profiling (16S rRNA sequencing) with urinary and fecal 1H NMR metabolomics, this study investigates early postnatal metabolic and microbial trajectories under supplementation exposure. Results highlight strong maturation-related succession in microbiome composition together with coherent functional shifts in fecal fermentation-related metabolites and time-structured remodelling in urinary profiles, supporting cross-layer triangulation while underscoring limitations typical of neonatal omics (small cohorts and high interindividual variability). Chapter 4 focuses on a longitudinal neonatal urinary metabolomics study aimed at assessing whether maternal SARS-CoV-2 infection at delivery, including persistence into early postpartum, is associated with deviations from expected postnatal metabolic trajectories. Mixed-effects modelling confirms that postpartum time is the dominant determinant of urinary profiles, whereas exposurerelated differences are modest and largely time-restricted, with limited baseline effects and selective group-by-time interactions superimposed on a strong maturation backbone. Moving beyond early-life applications, Chapter 5 presents a targeted systematic review of human Parkinson’s disease metabolomics (2019-2024). This chapter synthesizes converging biological themes, including alterations in energy metabolism, redox balance, lipid remodelling, and hostmicrobiome links, while emphasizing that translation remains constrained by methodological heterogeneity and insufficient validation. Finally, Chapter 6 provides the general conclusions and final considerations, highlighting the main contributions of the thesis, its current limitations, and the broader implications for neonatal metabolomics and translational research.