Longitudinal Gray Matter Trajectories in Pediatric Mild Traumatic Brain Injury

Abstract

Background and objectives: This prospective, longitudinal cohort study examined trajectories of brain gray matter macrostructure after pediatric mild traumatic brain injury (mTBI).

Methods: Children aged 8-16.99 years with mTBI or mild orthopedic injury (OI) were recruited from 5 pediatric emergency departments. Reliable change between preinjury and 1 month postinjury symptom ratings was used to classify mTBI with or without persistent symptoms. Children completed postacute (2-33 days) and/or chronic (3 or 6 months) postinjury T1-weighted MRI, from which macrostructural metrics were derived using automated segmentation. Linear mixed-effects models were used, with multiple comparisons correction.

Results: Groups (N = 623; 407 mTBI/216 OI; 59% male; age mean = 12.03, SD = 2.38 years) did not differ in total brain, white, or gray matter volumes or regional subcortical gray matter volumes. However, time postinjury, age at injury, and biological sex-moderated differences among symptom groups in cortical thickness of the angular gyrus, basal forebrain, calcarine cortex, gyrus rectus, medial and posterior orbital gyrus, and the subcallosal area all corrected p < 0.05. Gray matter macrostructural metrics did not differ between groups postacutely. However, cortical thinning emerged chronically after mTBI relative to OI in the angular gyrus in older children (d [95% confidence interval] = -0.61 [-1.15 to -0.08]); and in the basal forebrain (-0.47 [-0.94 to -0.01]), subcallosal area (-0.55 [-1.01 to -0.08]), and the posterior orbital gyrus (-0.55 [-1.02 to -0.08]) in females. Cortical thinning was demonstrated for frontal and occipital regions 3 months postinjury in males with mTBI with persistent symptoms vs without persistent symptoms (-0.80 [-1.55 to -0.05] to -0.83 [-1.56 to -0.10]) and 6 months postinjury in females and younger children with mTBI with persistent symptoms relative to mTBI without persistent symptoms and OI (-1.42 [-2.29 to -0.45] to -0.91 [-1.81 to -0.01]).

Discussion: These findings signal little diagnostic and prognostic utility of postacute gray matter macrostructure in pediatric mTBI. However, mTBI altered the typical course of cortical gray matter thinning up to 6 months postinjury, even after symptoms typically abate in most children. Collapsing across symptom status obscured the neurobiological heterogeneity of discrete clinical outcomes after pediatric mTBI. The results illustrate the need to examine neurobiology in relation to clinical outcomes and within a neurodevelopmental framework.

Figure 1. Summary of the Current Study Sample

Overall, of 3,075 eligible children, 967 consented to participate, and 846 returned for at least 1 assessment. The study aimed to have all children complete a postacute MRI and a second, chronic MRI at 3 or 6 months postinjury, with the timing of chronic scans randomly assigned at the postacute assessment. A total of 671 children completed at least 1 MRI. After excluding 141 (12.3%; 106 mTBI/35 OI) T1-weighted scans during quality checks, 1003 T1-weighted images were considered usable. An additional 8 (0.7%; mTBI/OI = 6/2) T1-weighted scans failed the ANTs processing pipeline. Thus, final analyses were based on data from 995 (87.0% if the original 1,144) T1-weighted scans in 623 children (407 mTBI/216 OI; see Table 1). Of the 541 children who completed a high-quality postacute scan, 372 (69%) also completed a chronic scan at 3 or 6 months postinjury (i.e., completed 2 high-quality scans; see eTable 2, links.lww.com/WNL/C922), and a subset completed only 1 high-quality scan (postacute/chronic = 169/82). ANTs = advanced normalization tools.

Figure 2. Summary of Linear Mixed-Effects Model Results for Regional Cortical Thickness With Robust (i.e., FDR Corrected) Group Differences

Three-way interactions between group, time postinjury, and age at injury (top) and between group, time, and sex (bottom) were observed for thickness in several cortical regions (indicated by color; see legend). Time and age moderated differences between the injury groups (i.e., TBI-OI; indicated by circles) for the angular gyrus and among the symptom groups based on parent (indicated by squares) and child (indicated by triangles) report for the calcarine cortex. Time and sex moderated differences between the injury groups for the posterior orbital gyrus, basal forebrain, and subcallosal area and among symptom groups based on parent report for the medial orbital gyrus, basal forebrain, calcarine cortex, gyrus rectus, and subcallosal area. Group comparisons are shown in Figure 3.

Figure 3. Group Differences in Regional Cortical Thickness Were Moderated by Time Postinjury, Age at Injury, and Biological Sex

Graphs illustrate robust differences in cortical thickness (mm) between the groups (i.e., Cohen d 95% confidence interval excluded 0), as denoted by *. For each effect, brain regions that showed the illustrated pattern between the groups are shown in the brain image insets for each graph (see Figure 2 legend).