Pilot longitudinal integrated transcriptomic-metabolomic study reveals immune and metabolic signatures in non-hospitalized healthcare workers with long COVID

<h4>Introduction</h4>Long COVID affects hundreds of millions of individuals worldwide, yet its underlying biological mechanisms remain incompletely understood, and the absence of validated biomarkers continues to limit diagnosis and clinical management. Most biomarker studies have focused on hospitalized patients with severe disease, leaving non-hospitalized populations, particularly healthcare workers, who are at high occupational risk, underrepresented. This gap may constrain the identification of biomarkers relevant to milder but persistent post-acute phenotypes.<h4>Methods</h4>We performed integrated transcriptomic and metabolomic profiling in a longitudinal cohort of non-hospitalized healthcare workers with long COVID (N = 12), primarily presenting with fatigue and brain fog, and matched controls who recovered from SARS-CoV-2 infection without sequelae (N = 35). Whole-blood RNA extracted from PAXgene tubes was profiled using the NanoString nCounter PanCancer Immune Panel. Serum metabolites collected pre- and post-infection were analyzed using untargeted ultra-high-performance liquid chromatography-mass spectrometry. Differential expression and metabolite abundance were assessed using linear models with false discovery rate correction. Significant features were integrated using network- and pathway-based approaches to identify coordinated immune-metabolic alterations in long COVID.<h4>Results</h4>Transcriptomic analysis identified 63 differentially expressed genes, including neutrophil-associated markers such as <i>S100A8</i> and <i>LY96</i>, consistent with activation of innate inflammatory pathways. Metabolomic profiling identified 24 annotated metabolites, with oxoglutarate exhibiting a distinct longitudinal trajectory, increasing in long COVID cases while decreasing in controls. Integrated network analysis highlighted central nodes (APP, RELA, ATF2, HLA-B) and revealed pathway-level convergence on necroptosis and serotonergic synapse signaling, suggesting coordinated immune-metabolic dysregulation rather than isolated gene-level effects.<h4>Discussion</h4>These findings generate hypotheses regarding potential links between persistent innate immune activation, metabolic reprogramming, and neurocognitive or systemic symptoms in long COVID. The observed signatures suggest immune-metabolic perturbations involving neutrophil-associated inflammatory pathways and broader cellular stress responses. However, given cohort size, platform-specific constraints, and cross-cohort heterogeneity, these signals should be interpreted at the pathway level and considered candidate mechanisms requiring validation in larger, independent cohorts of non-hospitalized individuals.