AbstractBackgroundBeyond the hypometabolic signature, indexed by [18F]FDG‐PET imaging, seen in the brain of Alzheimer’s disease (AD) patients, a transient brain glucose hypermetabolism has been recently identified in the early stages of AD. The cellular source of this dual metabolic response remains controversial, with glial cells gaining a lot of attention. In light of this, whether glial and synaptic fluid biomarkers are associated with [18F]FDG‐PET signal in the early stages of AD remains debatable. Thus, we evaluated the association of astrocyte, microglial and synaptic fluid biomarkers with [18F]FDG‐PET signal in an AD rat model.Method[18F]FDG‐PET imaging was conducted in ten‐month‐old (early amyloid stage) APP/PS1 (TgF344‐AD, n = 8‐17) and wild‐type (WT, n = 8‐15) rats. Glial (GFAP and sTREM2), synaptic (neurogranin), and amyloid biomarkers (Aß1‐40 and Aß1‐42) were quantified in the cerebrospinal fluid (CSF) and plasma through a multiplex immunoassay. T‐statistical maps of brain [18F]FDG‐PET and those fluid biomarkers were conducted at the voxel level using RMINC. Differences were considered statistically significant at p<0.05 (t>2).ResultA large hypermetabolic cluster was identified in ten‐month‐old TgF344‐AD (Fig. 1AB, local maxima, t(13) = 4.28). We found positive correlations between CSF TREM2 and brain [18F]FDG‐PET signal (Fig. 2AC, local maxima, t(13) = 4.12). In addition, we identified a positive correlation between plasma GFAP and brain [18F]FDG‐PET signal (Fig. 2BC, local maxima, t(13) = 10.62). Plasma TREM2, CSF GFAP, and plasma neurogranin did not associate with brain glucose metabolism.ConclusionOur findings suggest that CSF TREM2 and plasma GFAP are associated with brain metabolism before amyloid load reaches a plateau. Stronger association of plasma GFAP, rather than CSF GFAP, with brain pathology corroborates cross‐sectional human studies. One could argue that [18F]FDG‐PET hypermetabolism identified is likely a response to astrocyte and microglial changes.