Sex-specific differences in resting-state functional brain activity in pediatric concussion

Abstract

Pediatric concussion has a rising incidence and can lead to long-term symptoms in nearly 30% of children. Resting state functional magnetic resonance imaging (rs-fMRI) disturbances are a common pathological feature of pediatric concussion, though no studies have explicitly examined sex-differences with respect to this outcome, precluding a sex-specific understanding of the functional neuropathology of pediatric concussion. Therefore, we performed a secondary data analysis of rs-fMRI data collected on children with concussion (n = 29) recruited from in a pediatric hospital setting, with greater than 12:1 matched control data accessed from the open-source ABIDE-II database. Seed-based and region of interest (ROI) analyses were used to examine sex-based rs-fMRI differences; threshold-free cluster enhancement (TFCE) and a family-wise error (FWE) corrected p-values were used to identify significantly different clusters. In comparing females with concussion to healthy females, groupwise differences were observed irrespective of seed selected. Notably, we observed (in order of largest effect) hypo-connectivity between the anterior cingulate cortex of the salience network and the thalamus and precuneus (TFCE = 1473.5, p-FWE < 0.001) and the cingulate gyrus (TFCE = 769.3, p-FWE = 0.009), and the seed (posterior cingulate cortex (PCC)) of the default mode network and the paracingulate gyrus (TFCE = 1275.7, p-FWE < 0.001), occipital pole right (TFCE = 1045.0, p-FWE = 0.001), and sub-callosal cortex (TFCE = 844.9, p-FWE = 0.005). Hyper-connectivity was observed between the salience network seed and the cerebellum (TFCE = 1719.3, p-FWE < 0.001) and the PCC and the thalamus (TFCE = 1198.3, p-FWE < 0.001), cuneal cortex (1070.9, p-FWE = 0.001), and lateral occipital cortex left (TFCE = 832.8, p-FWE = 0.006). ROI analyses showed 10 and 5 significant clusters of hypo- and hyper-connectivity in females, respectively. Only one cluster of difference was found between males with concussion and healthy males on seed-based analyses, and 3 clusters on ROI analyses. There are alterations in rs-fMRI in females with concussion at one-month post-injury that are minimally present in males, which provides further evidence that recovery timelines in pediatric concussion may differ by sex.

Figure 1

rs-fMRI differences between children with concussion and their healthy peers (mixed-sex cohorts). Clusters (x, y, z) denote standard MNI coordinates at the center of cluster mass, and size represents number of voxels. Up to five clusters with the largest TFCE scores that survived a p-FWE < 0.05 per TFCE are displayed; see Supplemental Table 2 for all clusters that survived analysis.

Figure 2

Significantly increased (warm colours) and decreased (cool colours) ROI-to-ROI connectivity in children with concussion in comparison to controls (mixed-sex cohort).

Figure 3

rs-fMRI differences between males with concussion and healthy males. Clusters (x, y, z) denote standard MNI coordinates at the center of cluster mass, and size represents number of voxels. Only one cluster survived analysis.

Figure 4

Significantly decreased (cool colours) ROI-to-ROI connectivity in males with concussion in comparison to healthy male controls.

Figure 5

rs-fMRI differences between females with concussion and healthy females. Clusters (x, y, z) denote standard MNI coordinates at the center of cluster mass, and size represents number of voxels. Only one cluster survived analysis for the frontoparietal right; with multiple clusters identified for all other seeded networks.

Figure 6

Significantly increased (warm colours) and decreased (cool colours) ROI-to-ROI connectivity in females with concussion in comparison to healthy females.