BACKGROUND

Long COVID (LC) is characterised by persistent cognitive symptoms beyond 12 weeks after the acute phase, yet its biological underpinnings remain unclear. Sustained inflammation has been hypothesized as a potential underlying mechanism. This study investigated peripheral inflammatory markers and brain glucose metabolism in LC patients to explore underlying pathophysiological processes.

METHODS

This cross-sectional study included individuals from two LC cohorts in Argentina with persistent cognitive complaints and no prior neurological conditions. All participants underwent clinical evaluation, global cognition was screened using the Montreal Cognitive Assessment (MoCA), brain FDG-PET imaging, and blood sampling for peripheral inflammatory marker analysis. MoCA scores were normalised using UDS3 reference data. FDG-PET images were analysed using two independent pipelines (in-house and Neurocloud), each with separate healthy control databases and intensity normalisation strategies. Site-related differences were harmonised using ComBat. Regions showing consistent hyper- or hypometabolism across pipelines were selected. Peripheral blood levels of eotaxin, IFN-γ, IL-1β, IL-4, IL-6, IL-10, IL-17A, MCP-1, and TNF-α were quantified using multiplex ELISA in LC participants and unrelated healthy controls. Exploratory correlations were conducted between MoCA scores and cytokine levels.

RESULTS

Sixty-two LC participants were recruited (38 female, age 52 ± 12 years, MoCA z-score -0.5  ± 1.4), of which 49 had a FDG-PET scan. Evaluations were performed 20 ± 11.5 months post infection. Memory, concentration, fatigue, and attention complaints were reported by 87%, 91%, 69%, and 74% of participants, respectively. Compared with healthy controls, LC participants exhibited consistent hypermetabolism in bilateral frontal and parietal cortices (Table 1). Group comparisons revealed significantly increased levels of MCP-1 and IL-17A, and reduced levels of Eotaxin and IL-10 in LC vs controls (Figure 1). Exploratory correlations showed weak negative associations between MoCA scores and MCP-1 (r =0.25,p=0.06) and IL-17A (r =0.25,p=0.05).

CONCLUSIONS

Long COVID patients with persistent cognitive symptoms showed a pattern of frontal-dominant hypermetabolism and peripheral immune dysregulation. The reduced anti-inflammatory IL-10 and elevated pro-inflammatory markers support the hypothesis of immune dysregulation with a shift toward a pro-inflammatory state. Further studies are needed to determine whether these metabolic findings reflect chronic neuroinflammation, or a compensatory mechanism.