A common neural signature of brain injury in concussion and subconcussion

Abstract

The midbrain is biomechanically susceptible to force loading from repetitive subconcussive head impacts (RSHI), is a site of tauopathy in chronic traumatic encephalopathy (CTE), and regulates functions (e.g., eye movements) often disrupted in concussion. In a prospective longitudinal design, we demonstrate there are reductions in midbrain white matter integrity due to a single season of collegiate football, and that the amount of reduction in midbrain white matter integrity is related to the amount of rotational acceleration to which players' brains are exposed. We then replicate the observation of reduced midbrain white matter integrity in a retrospective cohort of individuals with frank concussion, and further show that variance in white matter integrity is correlated with levels of serum-based tau, a marker of blood-brain barrier disruption. These findings mean that noninvasive structural MRI of the midbrain is a succinct index of both clinically silent white matter injury as well as frank concussion.

Fig. 1. Visualization of spatial distribution of head hits in 38 collegiate football players (RSHI cohort) in a season of play.

(A) Plot of log count of hits by location/bin. Azimuth (longitude) and elevation (latitude) were binned into 10° square bins (36 bins for 360 azimuth and 18 bins for 180° elevation). At each location/bin, the total number of hits was counted across all 38 players in the RSHI cohort. Because the distribution of hit count was strongly right skewed (higher counts for fewer number of hits per location/bin), the log₁₀ of the count data was computed. The results are displayed in (A) as vectors, where the color and length of each vector scale by the log of the count of hits at that location. (B) Log count of hits by 10° × 10° azimuth and elevation bins. The histogram plots (y axis) the number of locations/bins at which (x axis) different numbers of hits were observed. Coloring on the histogram serves as a color scale for the data displayed in (A). (C) Histogram of distribution of linear acceleration for all hits. (D) Histogram of distribution of rotational acceleration for all hits. An interactive model of the data shown in (A) can be found at www.openbrainproject.org/tbi.

Fig. 2. White matter integrity is reduced postseason compared to preseason in the RSHI cohort.

FA in the right CST midbrain ROI was significantly reduced postseason compared to preseason. The red data points/lines correspond to the two players (of the group of 38) who sustained a frank concussion.

Fig. 3. Correlation between head hits with changes in FA.

In all plots, black circles indicate individuals without clinically diagnosed mTBI and red circles indicate the two individuals who suffered a concussion between the pre- and postseason MRI. (A) The scatter plot shows the relation between changes in FA in the right midbrain and the number of head impacts with rotational acceleration equal to or greater than 1 SD above the group mean. The direction of the relation indicates that more trauma is associated with greater reductions in the structural integrity of white matter. This relation remained when excluding the two participants who suffered a clinically defined concussion (r = −0.42, P < 0.012). (B) The relation between rotational acceleration and changes in FA holds when controlling for linear acceleration by restricting the analysis to hits that exceed the threshold for rotational acceleration but do not exceed the threshold for linear acceleration. This relation remained when excluding the two participants who suffered a clinically defined concussion (r = −0.44, P < 0.008). (C) The number of head hits with linear acceleration greater than 1 SD above the mean is negatively correlated with the change (postseason minus preseason) in FA in the midbrain. This relation, however, was not significant when excluding the two participants who suffered a clinically defined concussion (r = −0.32, P = 0.06). (D) The relation between linear acceleration and changes in FA goes away when controlling for rotational acceleration by restricting the analysis to hits that exceed the threshold for linear acceleration but do not exceed the threshold for rotational acceleration. This lack of a relation between linear acceleration and changes in DTI remained absent when excluding the two participants who suffered a clinically defined concussion (r = −0.11, P = 0.51).

Fig. 4. Control analyses ensured that the relation between rotational acceleration and changes in FA does not depend on thresholds.

The plot shows (black stars) the rotational acceleration values corresponding to decile binning of the data. For each decile, the correlation between the number of hits (at that threshold or higher) and changes in FA was computed (filled red circles). Regardless of the threshold used, there was a significant correlation between changes in FA and number of head hits.

Fig. 5. White matter structural integrity of the midbrain indexes clinically defined mTBI and relates to peripheral tau.

(A) ROI analyses of the right CST ROI in the midbrain show significantly reduced FA in individuals diagnosed with a concussion compared to matched controls. (B) There is a negative correlation between peripheral tau and structural integrity of the midbrain in concussed individuals, indicating that higher levels of peripheral tau are associated with greater reductions in structural integrity of the midbrain CSTs.

Fig. 6. SVR analysis relating the spatial fingerprint of head impacts to hemispheric asymmetries in white matter.

Each player has a “spatial fingerprint” of the distribution of hits around the head over the entire season. That pattern was used to train a linear SVR model to predict variance across participants in hemispheric asymmetries in white matter changes. (A) Visualization of relevance of hit location to SVR analysis. The images are a graphical representation of the SVR model relating variance across participants in the spatial distribution of hits to variance across participants in the laterality of white matter changes. This visualization of the SVR analysis was computed by correlating the variance across participants in feature weights (at each location/bin) with the variance across participants in hemispheric asymmetries in white matter changes. The results are plotted as vectors that are scaled in length and color by variance explained (Bonferroni thresholded at 05/648 = 0.00007, or a critical r² of 0.36). (B) SVR analysis of hemispheric asymmetry and comparison. Using an n − 1 cross-validation approach, a linear SVR model captures 60% of the variance in hemispheric asymmetries in white matter changes. The variance explained by this analysis differs significantly from chance, as shown through permutation testing (100,000 shuffles). An interactive model of the data shown in (A) can be found at www.openbrainproject.org/tbi. FDR, false discovery rate.