Point-of-care Diagnostic Framework for Fibromyalgia Using Integrated Vibrational Spectroscopy and Metabolomics

Background There is a critical need for objective diagnostic strategies for syndromes that rely on subjective questionnaires to ensure accurate and reliable diagnosis. Fibromyalgia (FM), one of the most common rheumatic disorders, remains particularly challenging to diagnose because of symptom overlap with related conditions, especially rheumatoid arthritis (RA). This exploratory study evaluated the feasibility of a combined spectroscopic–metabolomic workflow for distinguishing FM from RA and healthy controls (HC). Methods Whole blood specimens were analyzed from 40 patients with FM, 20 patients with RA, and 10 HC participants. The analytical workflow combined portable Fourier-transform infrared (FTIR) spectroscopic fingerprinting with mass spectrometry (MS)-based metabolite identification. Potential confounding factors affecting analytical signatures were systematically evaluated, and alternative extraction protocols were compared. Methanol (MeOH) and methanol/1-butanol (MeOH/BuOH) extraction methods gave a good metabolome coverage and were selected for subsequent analyses. Soft independent modeling by class analogy (SIMCA) and partial least squares discriminant analysis (PLS-DA) were used for classification, while partial least squares regression (PLSR) was used to correlate FTIR spectral features with biologically relevant metabolites identified by MS. Results SIMCA models demonstrated good classification between FM and HC, with interclass distances (ICD) exceeding 4.1 for MeOH extraction and 4.6 for MeOH/BuOH extraction. MS-based metabolomic analyses identified oligopeptides, inosine monophosphate, and signaling lipid molecules as major contributors to group differentiation, suggesting probable dysregulation of oxidative stress and inflammatory signaling pathways, as well as purine, amino acid, and free fatty acid metabolism. PLSR models showed strong correlations between FTIR spectral data and MS intensities, including inosine monophosphate, N-acylethanolamines (NAE), monoacylglycerols (MAG), N-fructosyl phenylalanine, N-fructosyl isoleucine, Ser-Phe, and N-acetylhistidylprolinamide (R ≥ 0.79; SECV ≤ 0.30). Conclusions These findings demonstrate the potential of integrating rapid FTIR spectroscopic fingerprinting with MS-driven metabolomics to identify biologically relevant signatures associated with fibromyalgia. The strong classification performance and metabolite correlations support the potential translation of this diagnostic pipeline into a rapid, point-of-care approach for objective FM diagnosis.