The brain's "dark energy" puzzle: How strongly is glucose metabolism linked to resting-state brain activity?
Tommaso Volpi,Erica Silvestri,Marco Aiello,John J Lee,Andrei G Vlassenko,Manu S Goyal,Maurizio Corbetta,Alessandra Bertoldo
DOI: https://doi.org/10.1177/0271678x241237974
2024-03-07
Journal of Cerebral Blood Flow & Metabolism
Abstract:Journal of Cerebral Blood Flow &Metabolism, Ahead of Print. Brain glucose metabolism, which can be investigated at the macroscale level with [18F]FDG PET, displays significant regional variability for reasons that remain unclear. Some of the functional drivers behind this heterogeneity may be captured by resting-state functional magnetic resonance imaging (rs-fMRI). However, the full extent to which an fMRI-based description of the brain's spontaneous activity can describe local metabolism is unknown. Here, using two multimodal datasets of healthy participants, we built a multivariable multilevel model of functional-metabolic associations, assessing multiple functional features, describing the 1) rs-fMRI signal, 2) hemodynamic response, 3) static and 4) time-varying functional connectivity, as predictors of the human brain's metabolic architecture. The full model was trained on one dataset and tested on the other to assess its reproducibility. We found that functional-metabolic spatial coupling is nonlinear and heterogeneous across the brain, and that local measures of rs-fMRI activity and synchrony are more tightly coupled to local metabolism. In the testing dataset, the degree of functional-metabolic spatial coupling was also related to peripheral metabolism. Overall, although a significant proportion of regional metabolic variability can be described by measures of spontaneous activity, additional efforts are needed to explain the remaining variance in the brain's 'dark energy'.
neurosciences,endocrinology & metabolism,hematology