Local activity determines functional connectivity in the resting human brain: a simultaneous FDG-PET/fMRI study

Abstract

Over the last decade, synchronized resting-state fluctuations of blood oxygenation level-dependent (BOLD) signals between remote brain areas [so-called BOLD resting-state functional connectivity (rs-FC)] have gained enormous relevance in systems and clinical neuroscience. However, the neural underpinnings of rs-FC are still incompletely understood. Using simultaneous positron emission tomography/magnetic resonance imaging we here directly investigated the relationship between rs-FC and local neuronal activity in humans. Computational models suggest a mechanistic link between the dynamics of local neuronal activity and the functional coupling among distributed brain regions. Therefore, we hypothesized that the local activity (LA) of a region at rest determines its rs-FC. To test this hypothesis, we simultaneously measured both LA (glucose metabolism) and rs-FC (via synchronized BOLD fluctuations) during conditions of eyes closed or eyes open. During eyes open, LA increased in the visual system, and the salience network (i.e., cingulate and insular cortices) and the pattern of elevated LA coincided almost exactly with the spatial pattern of increased rs-FC. Specifically, the voxelwise regional profile of LA in these areas strongly correlated with the regional pattern of rs-FC among the same regions (e.g., LA in primary visual cortex accounts for ∼ 50%, and LA in anterior cingulate accounts for ∼ 20% of rs-FC with the visual system). These data provide the first direct evidence in humans that local neuronal activity determines BOLD FC at rest. Beyond its relevance for the neuronal basis of coherent BOLD signal fluctuations, our procedure may translate into clinical research particularly to investigate potentially aberrant links between local dynamics and remote functional coupling in patients with neuropsychiatric disorders.

Keywords: PET/MR imaging; energy metabolism; functional connectivity.

Figures

Figure 1.

LA and FC during resting-state CLOSED (left) and OPEN (right) conditions. A, SPMs of whole-brain voxelwise one sample t tests for FDG uptake controlling for each subject's grand mean signal. B, SPMs of whole-brain voxelwise one-sample t tests on individual FC maps revealing rs-FC for seed ROI V1 (V1-FC). Regions (red) with V1-FC during the CLOSED condition include primary and secondary visual cortex, and inferior parietal and temporoparietal areas. Additional regions with V1-FC during the OPEN condition include cingulate, insular, and prefrontal cortices; brainstem; thalamus; and cerebellum. C, SPMs of whole-brain voxelwise one sample t tests on individual FC maps revealing rs-FC for averaged seed ROI of the SAL network (SAL-FC). Regions (red) with SAL-FC during the CLOSED condition include cingulate, insular, superior parietal, and prefrontal cortices; brainstem; thalamus; and cerebellum. Additional regions with SAL-FC during the OPEN condition included primary and secondary visual, and anterior cingulate cortices. All SPMs are thresholded at an FWE-corrected p value of <0.05 and are superimposed on a high-resolution T1-weighted anatomical image. x, z, Coordinates of brain slices in MNI space; R, right side of the brain.

Figure 2.

Quantitative relationship between LA and FC in the resting human brain. A, SPMs of whole-brain voxelwise ANCOVA for LA changes controlling for each subject's grand mean signal. Regions (cyan) with increased LA during the OPEN condition included V1, aI, aCC, and mCC. FWE corrected, p < 0.05. R, Right. B, SPMs of whole-brain voxelwise independent-samples t tests on individual FC maps for seed ROIs (cyan) V1 (V1-FC; top row) and SAL hubs (SAL-FC; bottom row) revealing FC changes. Regions (yellow) with increased V1-FC during the OPEN condition include SAL hubs, thalamus, brainstem, and cerebellum; the main region with increased SAL-FC during the OPEN condition was V1 and parts of aCC. p < 0.05, FWE corrected. C, Voxelwise conjunction showing regions (violet) with increased LA and FC. An exemplary distribution of voxel values of LA (y-axis) and V1-FC (x-axis) of aI taken from the first subject is shown on the right side. To quantify the relationship between LA and FA, we calculated the SpaSi[LA, FC] ROI coefficient for each subject and each ROI as the voxelwise spatial correlation between normalized LA and between-network FC values, and subsequently analyzed them in a two-way ANOVA. D, Post hoc independent-samples t tests revealed significantly increased SpaSiOPEN for ROIs V1, aCC, and aI. **p < 0.005, ***p < 0.0005. Error bars denote 5th and 95th percentiles. No difference occurred in V1 for within-network connectivity (V1[V1-FC]) and in mCC for between-network FC. E, Mean square of SpaSiCLOSED and SpaSiOPEN, indicating the voxelwise variability of FC explained by LA. *p < 0.05, ***p < 0.0005, one-sample t tests performed separately for SpaSiCLOSED and SpaSiOPEN. Cluster coordinates for all SPMs and cluster sizes for conjunction regions can be found in Table 1.