Persistent Disruption of Brain Connectivity after Sports-Related Concussion in a Female Athlete

Abstract

Structural and functional connectivity (FC) after sports-related concussion (SRC) may remain altered in adolescent athletes despite symptom resolution. Little is known, however, about how alterations in structural connectivity and FC co-present in female athletes whose symptom recovery tends to be prolonged. Despite resolution of symptoms, one month after her second SRC, an 18-year-old female athlete had decreased structural connectivity in the corpus callosum and cingulum, with altered FC near those regions, compared with other SRC and orthopedically injured athletes. Findings show persistent effects of SRC on advanced brain imaging and the possibility of greater vulnerability of white matter tracts in females.

Introduction

Imaging studies of sports-related concussion (SRC) in adolescent athletes have found that structural connectivity measured by diffusion tensor imaging (DTI) and functional connectivity (FC) measured from resting state functional magnetic resonance imaging (rs-fMRI) differed from non-concussed athlete controls.^1,2 In particular, these brain connectivity changes are observed in athletes even after symptoms have resolved and they are medically cleared to return-to-play (RTP).^3,4 There is also recent evidence that exposure to playing a contact or collision sport without incurring an SRC is sufficient to produce changes on DTI,^5,6 implicating the cumulative effects of repetitive subconcussive head impacts on the organization of white matter (WM) tracts.

Epidemiological and observational studies of SRC in student athletes have reported a higher rate of concussion and more persistent post-concussion symptoms (PCS) in female than male adolescent athletes playing the same sport.^7,8 Koerte and associates⁵ (2017) also found that pre- and post-season changes in WM tracts on DTI for adolescent hockey players without concussion were specific to females, suggesting greater vulnerability to diffuse axonal injury (DAI).

The neural underpinnings of sex differences in SRC risk and WM pathology have been informed recently by translational research demonstrating that axonal architecture is more robust in male than female rats and confirmed by corresponding findings in human axons.⁹ Dolle and colleagues⁹ (2018) found that human males had 80% greater axonal cross-sectional area and 55% greater microtubules than females, which may have predisposed the female microtubules to break more easily, resulting in more undulations of the axon and disruption of axonal transport after mild traumatic brain injury (mTBI).⁹ Human neuroimaging data to support the slower recovery from SRC, and the more adverse effects of repetitive subconcussive head impacts in adolescent female athletes, are sparse, however.⁷ The recent Concussion in Sport Consensus Guidelines¹⁰ also acknowledged that group studies may obscure variability in outcomes,¹¹ which provided the impetus for the current report.

We report the deidentified case of Jane—a female adolescent athlete who had extensive exposure to soccer and sustained two SRCs, separated by three years, before DTI and rs-fMRI analyses. We hypothesized more pronounced alterations in both types of connectivity in Jane relative to both an orthopedic injury (OI) comparison group or to other concussed athletes who had no SRCs before the current study.

Patient and Methods

All methods were in full compliance with the Institutional Review Boards (IRB) at Baylor College of Medicine and UTHealth McGovern Medical School and followed the principles of the Declaration of Helsinki. All participants provided assent; parents or legal guardians also provided written consent and acknowledged that all data were fully anonymized and cannot be identified via the article.

Participants

History

This right-handed, typically developed female student had no history of neurological or psychiatric disorder before her first SRC (SRC1), but at the time of her second SRC (SRC2), she had a 15-year history of playing soccer. Jane sustained SRC1 in March 2012 at age 15 when her head hit the ground during a high school soccer game. She reported acute headache, confusion, dizziness, fatigue, diplopia, and photophobia without loss of consciousness and sought treatment at an emergency department five days post-injury.

Jane was first seen in the concussion clinic three weeks post-injury and had homebound instruction for three months. She was treated with Elavil^® (amitriptyline, Qualitest Pharmaceuticals/Vintage Pharmaceuticals, Huntsville, AL) at five weeks, but did not tolerate it; her physician switched her to Topamax^® (topiramate, Janssen Pharmaceuticals, Raritan, NJ). At two months post-SRC1, she started amantadine (Endo, Ballsbridge, Dublin, Ireland) because of concentration problems. She was asymptomatic by late June, released to resume exertional activity, and cleared for RTP by her clinical provider in July 2012, nearly five months post-injury.

At age 18, Jane sustained SRC2 in January 2015 when her head hit the ground during a soccer game. Similar to SRC1, she reported no loss of consciousness, and she denied post-traumatic amnesia (PTA). When examined in clinic on day 3, she stated that the mild headache and balance difficulties that had been present on the day of injury had cleared. ImPACT composite scores on Verbal Memory and Visual Memory, however, were below her pre-season baseline. The RTP protocol was initiated on day 4. On day 7, she returned to clinic and complained of balance problems and dizziness; her ImPACT Visual Memory Composite was still below baseline. On post-SRC2 day 10, Jane was asymptomatic, her ImPACT composite scores had returned to baseline, and she was cleared to RTP.

Having met the eligibility criteria for enrollment in a study that was ongoing, including the criterion for undergoing brain imaging (asymptomatic status and clearance for RTP by day 30), she had DTI and MRI on day 30. As an ad hoc analysis, FC was measured from her rs-fMRI data in a limited number of regions located near WM structures that were aberrant in the DTI analysis.

Comparison groups

We compared Jane's imaging findings with data collected from two groups of participants in the ongoing study, including 11 concussed adolescents (SRC comparison group) who were closest in age to Jane, had played a contact sport, and had also been cleared to RTP by day 30, and were imaged (mean age at injury = 16.41 years, standard deviation [SD] = 0.98, range 15–18 years, n = 10 males, one female). Nine adolescent contact athletes who had sustained a mild OI (Abbreviated Injury Scale ≤2; www.aaam.org/abbreviated-injury-scale-ais/ ; accessed November 6, 2018) (age: mean = 16.64, ±1.02, n = four male, five female) were in the OI comparison group.

One of the contact athletes in the OI comparison group reported a previous concussion, whereas none of the athletes in the concussed comparison group reported a previous concussion. The number of seasons that the athletes in the concussed and OI comparison groups had played a contact sport and the age when each athlete was first exposed to a contact sport were not available. Although the post-injury interval for imaging was longer in the OI group, (SRC: mean = 33.1 days ±1.7; OI: mean = 124.7 days ±69.0), we have found in previous studies that imaging findings are stable between the first week and three months after OI.¹²

Because Jane fell asleep during the resting state acquisition, we were unable to use her resting state sequence. An additional resting state sequence, however, testing a different protocol was acquired in a subset of participants (n = 6 SRC, 6 OI), and Jane was fully awake during this acquisition. Participants not scanned with this sequence were excluded from rs-fMRI analysis.

Procedure

During the resting state acquisition, which immediately preceded DTI in the same imaging session, the MRI technologist instructed the participants to lie still and to keep their eyes fixated on a red circle in their line of sight. Immediately after the rs-fMRI sequence, participants were queried to assess their wakefulness, and none reported sleepiness or had fallen asleep.

Image data acquisition

Whole brain imaging was performed using a 32-channel head coil on a Siemens 3T Magneton system (Siemens Medical Solutions, Erlangen, Germany) at the Center for Advanced Magnetic Resonance Imaging, Baylor College of Medicine, Houston, TX. Regular quality assurance testing was performed throughout the study, and no issues were detected. Transverse multi-slice spin echo, single shot, echo planar imaging (EPI) sequences were applied (9000 msec repetition time [TR], 85 msec echo time [TE], 2.0 mm slices, with no gap, 73 slices) with iPAT3 for DTI acquisition. Diffusivities were measured along 64 directions using a low b-value of 0, and high b-value of 1000 sec/mm². Eddy current distortion and head motion artifact were corrected.

For fMRI, blood oxygen level dependent (BOLD) T2*-weighted echo planar (EPI) were acquired as 200 volumes with 48 axial slices of 3.3- mm thickness with a 0 mm gap, using a 212-mm field of view (FOV), 64 × 64 matrix, TR of 3000 msec, TE of 30 msec, and an 80-degree flip angle. A set of three dimensional high-resolution T1-weighted images were also acquired in 176 sagittal slices of 1.2 mm thickness (no gap) with 240 mm FOV, 240 × 256 matrix, TR of 2300 ms, TE of 2.96 ms, and a 9.0-degree flip angle. Additional anatomical sequences were acquired to evaluate neuropathology. A senior neuroradiologist (JVH) reviewed and coded the structural MRI findings without knowledge of group membership.

DTI analysis

Consistent with meta-analyses reporting the WM tracts most frequently affected by SRC and non-sports mTBI,¹³ we measured fractional anisotropy (FA) and mean diffusivity (MD) for the genu of the corpus callosum, the total corpus callosum, the left and right sides of the uncinate fasciculus, and the cingulum bundle. The genu of the corpus callosum was also reported to be especially vulnerable to SRC in female athletes.^14,15 Shear and eddy current distortion, as well as head motion artifact, were corrected by using the Philips' Research Image Development Environment (PRIDE^®) registration tool. Regions of interest of the corpus callosum were traced manually on the midsagittal plane, following a published protocol.¹⁶

The automated Philips three-dimensional fiber tracking 4.1v3 Beta 2 program was utilized to examine fiber tracks passing through the corpus callosum, the left and right sides of the uncinate fasciculus, and the cingulum bundle.¹⁶ The algorithm for fiber tracking was based on the fiber assignment by continuous tracking method.¹⁷ If the FA in the voxels was <0.2, or if the angle between adjacent voxels was >7 degrees, fiber tracking terminated. To evaluate interrater reliability, DTI analysis of the subregions was performed for six randomly selected participants (n = 3 concussed, 3 OI controls) by two experienced raters. To ensure intrarater reliability, each region of interest was analyzed independently by each rater twice. Shrout-Fleiss intraclass correlation coefficients were >0.90 for intra- and interrater reliability.¹⁸

fMRI image processing and analysis

The FC image processing and analyses were conducted using the Functional Connectivity Toolbox (CONN)¹⁹ within SPM8 (Wellcome Department of Cognitive Neurology, University College, London, UK), implemented in Matlab (Mathworks Inc., Sherborn, MA), as reported in Newsome and coworkers.¹ Seeds to measure the integrity of FC around the genu and total corpus callosum were located in the anterior, posterior, and subcallosal gyri. Significance was defined by voxel (height) threshold p < 0.001,²⁰ and cluster threshold (p < 0.05) false discovery rate (FDR) corrected for multiple comparisons across the whole brain. Bonferroni correction for three seeds for both tails of two between-group t tests was applied (p = 0.05 / [3 × 2 × 2] = 0.0042).

Results

Radiological reading of MRI

Clinical review of Jane's MRI disclosed no abnormalities. Similarly, there were no anatomical abnormalities in the MRIs of the SRC and OI comparison groups.

DTI findings

Figure 1 shows that Jane's FA of 0.451 for the total corpus callosum was below the range of the nine athletes with OI (mean: 0.479, range: 0.466 – 0.506), 95% confidence interval [CI] (0.470 – 0.489), and the 11 athletes who sustained SRC (mean: 0.485, range 0.456 – 0.513), 95% CI (0.475 – 0.496). Table 1 shows that Jane's FA for the genu of the corpus callosum and both the right and left cingulum bundle was lower than the FA distributions of the OI and SRC comparison groups.

FIG. 1.

Boxplots representing fractional anisotropy (FA) measured from Jane's total corpus callosum and the distributions of FA from the concussed and orthopedic injury (OI) comparison groups. The dot in the center of each boxplot indicates the mean, the upper horizontal line indicates the upper quartile (75th percentile) value, and the lower horizontal line indicates the lower quartile (25th percentile). The vertical line extending above the box indicates the highest value, whereas the line extending below the box indicates the lowest value. TBI, traumatic brain injury; MD, mean diffusivity.

Table 1.

Mean Diffusivity and Fractional Anisotropy Values with Associated Standard Deviations and Ranges for Jane and the Sports-Related Concussion and Orthopedic Injury Groups

	MD (SD) range			FA (SD) range
Brain region of interest	Jane	SRC	OI	Jane	SRC	OI
Genu	0.837	0.883	0.852	0.418*	0.459	0.461
		(0.033)	(0.026)		(0.016)	(0.017)
		0.819–0.933	0.802–0.888		0.428–0.480	0.440–0.493
CCTOT	0.861	0.895	0.890	0.451*	0.485	0.479
		(0.035)	(0.022)		(0.016)	(0.013)
		0.850–0.967	0.854–0.917		0.456–0.513	0.466–0.506
L Cing	0.803*	0.758	0.761	0.395*	0.470	0.451
		(0.030)	(0.030)		(0.026)	(0.018)
		0.700–0.785	0.720–0.808		0.434–0.506	0.421–0.484
R Cing	0.785	0.754	0.759	0.393*	0.447	0.425
		(0.027)	(0.032)		(0.025)	(0.021)
		0.713–0.794	0.717–0.792		0.408–0.487	0.400–0.456
L Unc#		0.812	0.797	0.390	0.401	0.392
		(0.019)	(0.033)		(0.027)	(0.025)
		0.776–0.842	0.758–0.850		0.343–0.431	0.353–0.433
R Unc	0.815	0.814	0.802	0.367	0.403	0.394
		(0.022)	(0.026)		(0.027)	(0.031)
		0.774–0.853	0.766–0.840		0.340–0.432	0.354–0.451

MD, mean diffusivity; FA, fractional anisotropy; SD, standard deviation; SRC, sports-related concussion; OI, orthopedic injury; CC_TOT, corpus callosum total; L Cing, left cingulate; R Cing, right cingulate; L Unc, left uncinate; R Unc, right uncinate.

^#Jane's left uncinate FA was unmeasurable because of a technical problem.

^*Jane's DTI metric falls outside the range of one or both comparison groups.

The above differences in FA and MD were present despite Jane being within two months of the oldest athlete in the combined SRC and OI groups. Consistent with ongoing WM maturation in adolescents, FA tends to increase with age.²¹ Jane's MD measured from the left cingulum bundle was toward the high end of the MD distributions of the comparison groups, but her MDs for the other tracts were within range of the other athletes. A sagittal view of Jane's DTI streamlines represents a loss of fibers in the right parietal region compared with a control athlete who was two years younger (Fig. 2).

FIG. 2.

(A) Sagittal view of diffuse tensor imaging (DTI) streamlines showing a reduction of fibers in Jane's right parietal region at 30 days after her second sports-related concussion, at age 18 years when she had recovered clinically and was cleared to return to play, as compared with (B), sagittal view of DTI of a 16-year-old female athlete who was in the orthopedic injury control group.

Functional connectivity

Because structural connectivity of Jane's total corpus callosum was aberrant, we measured FC of the posterior cingulate (PCC), anterior cingulate (ACC), and the subcallosal cortex (SCC), which are anatomically contiguous to the corpus callosum (Table 2 and Fig. 3).

Table 2.

Functional Connectivity t-Tests between Sports-Related Concussion Participant Jane and Other Sports-Related Concussion Participants and between Jane and Orthopedic Injury Participants

Region of interest Cluster-Level	p Value (corrected)a	Cluster Size (k)b	Most Significant Coordinates (x y z)c	Locationd
a. Posterior cingulate cortex (PCC)
OI > Jane	NS
Jane > OI	0.000103	95	−8 −56 −36	Cerebel 8 L, 9 R; Vermis 9 d
SRC > Jane	NS
Jane > SRC	NS
b. Anterior cingulate cortex (ACC)
OI > Jane	0.000081	85	42 14 4	IFG (oper) R, Ins R
	0.001207	55	32 2 66	MFG R, PreCGR, SFG R
Jane > OI	0.0000001	352	56 −12 −12	aMTG R, pMTG R, aSTG R
	0.000001	131	0 62 10	FP L&R
	0.000019	97	−50 −8 −20	aMTG L, aSTG L
	0.000040	87	6 −54 6	Precun, LG R
	0.000209	69	−12 −46 −2	LG L, Cerebel 45 L
	0.000209	68	26 −46 −22	ToFus R, Cerebel 45 R, Cerebel 6 R
	0.000359	62	22 −22 −16	Hippocampus R
	0.003380	42	6 −46 28	Posterior cingulate
	0.003380	42	8 −30 4	Thalamus L
Jane FC>SRC	NS
Jane FC<SRC	NS
c. Subcallosal cortex (SCC)
OI > Jane	NS
Jane > OI	0.001080	68	−34 −12 64	PreCG L
SRC > Jane	NS
Jane > SRC	0.000036	99	14 −90 26	Cun R, sLOC R, OP R
	0.000883	62	6 12 34	ACC
	0.001903	52	14 −64 −4	LG R

NS = not significant; a, anterior; cerebel, cerebellum; FP, frontal pole; IFG, inferior frontal gyrus; ins, insula; L, left; R, right; LG, lingual gyrus; MFG, middle frontal gyrus; MTG, middle temporal gyrus; oper, operculum; OI, orthopedic injury; OP = occipital pole; p, posterior; PCC, superior frontal gyrus; Precun, precuneus; SRC, sports-related concussion; STG, superior temporal gyrus; ToFus, temporal occipital fusiform cortex; Cun, cuneus; sLOC, superior lateral occipital cortex; ACC, anterior cingulate cortex

^aProbability at the cluster level of significance after random field theory family-wise error correction over the whole brain search volume. Cluster probability also survives Bonferroni correction for four seeds, two comparison groups, and two directions (p = 0.05 / [3 × 2 × 2] = 0.004).

^bNumber of voxels within a cluster.

^cNegative values along the x-axis are defined to be in the participant's left hemisphere.

^dThe number after “Cerebel” and “Vermis” denotes the region number.

FIG. 3.

Imaging 30 days after her second sports-related concussion (SRC2) when Jane was asymptomatic and had been cleared to return-to-play (RTP) disclosed that she had greater functional connectivity, most often positive, compared with SRC and orthopedic injury (OI) adolescent athletes. (A) Panel depicting increased functional connectivity in Jane compared with the other SRC athletes between a seed in subcallosal cortex and three clusters. (B) Panel depicting increased FC in Jane compared with orthopedic injury (OI) athletes between a seed in the anterior cingulate (AC) and seven clusters throughout the brain. R, right; L, left; Cun, cuneus; AC, anterior cingulate; LG, lingual gyrus; MTG, middle temporal gyrus; FP, frontal pole; PreCun, precuneus; ToFus, temporal occipital fusiform cortex; Hipp, hippocampus.

PCC

Relative to the OI group, Jane demonstrated an increased positive relation between the PCC seed and a cluster incorporating three regions of the cerebellum (left cerebellum, right cerebellum, and vermis; beta = 0.44). Jane had no regions with decreased FC of the PCC, and no differences relative to the SRC group.

ACC

Relative to the OI group, Jane demonstrated increased positive relation between the ACC seed and nine clusters throughout the brain consisting of the bilateral medial frontal pole (beta = 0.72), right middle and superior temporal gyri (beta = 0.69), right temporo-occipital fusiform gyrus (beta = 0.47), PCC (beta = 0.53), precuneus and right lingual gyrus (beta = 0.63), right hippocampus (beta = 0.60), left thalamus (beta = 0.59), and left lingual gyrus and cerebellum (beta = 0.47). She also showed significantly decreased FC between the ACC and right inferior frontal gyrus (pars operculum) and right insula (beta = -0.52), and right precentral, middle, and superior frontal gyri (beta = -0.49), but no differences from the SRC group.

SCC

Relative to the OI comparison group, Jane demonstrated increased positive FC between the subcallosal seed and left precentral gyrus (beta = 0.39). Jane had no regions with decreased FC of the PCC. As seen in Figure 3, in comparison with the SRC comparison group, Jane's FC from SCC was increased in the positive direction in the ACC (beta = 0.57), right lingual gyrus (beta = 0.33), and right cuneus, lateral occipital (superior) cortex, and occipital pole (beta = 0.41). There were no regions with which she had decreased subcallosal FC compared with the SRC group.

Discussion

Despite clinical recovery and having been cleared for RTP on day 10 after SRC2, this asymptomatic athlete, whose cognitive performance recovered to baseline, had aberrant structural and functional connectivity when imaged on day 30. The FA measured from the genu, total corpus callosum, and both the left and right cingulum bundle was below the distributions of other clinically recovered, concussed adolescents and/or a group of non-concussed adolescents who had sustained an OI.

Considering that Jane was older than all but one of the athletes in the comparison groups and that FA increases with age in typically developing adolescents,²¹ her DTI findings were aberrant. Reduced FA suggests less anisotropic diffusion of water parallel to the long axis of these compromised tracts. The MD in Jane's left cingulum bundle was correspondingly high on day 30, exceeding the values of all but one of the other athletes. These DTI findings are consistent with reduced microstructural integrity of major WM tracts, ostensibly a biomarker for DAI.⁹

Jane's FC from areas proximal to her structurally altered WM differed from both comparison groups. Relative to the SRC comparison group, Jane's FC was increased between the SCC seed and bilateral ACC and right occipital regions. Such an increase may have been compensatory and contributed to Jane's cognitive recovery from SRC2. Potentially, impaired diffusion along axons connecting frontal and occipital regions (cingulum bundles and corpus callosum) may have permitted less precise neural transmission, allowing greater signal in gray matter regions that would not have been targeted otherwise. Increased activation in posterior areas of the brain in adolescents with moderate-to-severe TBI was partially attributed to decreased cerebral blood flow (CBF) in frontal brain regions.²² Altered CBF may also account for aberrant brain connectivity after SRC as seen in Jane's data.⁴

Although imaging was not acquired until day 30 after SRC2, we acknowledge the possibility that the aberrant structural and functional connectivity may represent residual effects from Jane's first SRC that resulted in persistent PCS⁴ and dissociation between meeting criteria for RTP within 10 days after SRC2 and her aberrant brain connectivity.^4,5,23 The DTI results implicating persistent WM injury are also consistent with recent imaging studies showing disrupted functional and/or structural connectivity in student athletes who had been cleared for RTP.^1,24–26 Moreover, recent longitudinal DTI findings in concussed high school and college athletes showed no evidence of normalization after six months.²⁷

In summary, Jane's aberrant brain connectivity on advanced brain imaging is a striking example of neurobiological dysfunction outlasting clinical recovery based on resolution of symptoms and cognitive performance.¹¹

Cumulative effects of playing soccer for 15 years during a period of dynamic brain maturation may also have contributed to Jane's aberrant structural and FC.⁵ Recent imaging studies of student athletes participating in collision or contact sports have shown changes in DTI and FC between pre-season and post-season scans.^4,5 Koerte and colleagues⁵ (2017) also found that post-season DTI changes from baseline were present in non-concussed female—but not male—hockey players. Consistent with this sex-specific finding, a translational study found that post-concussion undulations of the axon and breakage of microtubules were greater in the brains of female rats than in the males.⁹

Correspondingly, DTI disclosed more severe WM injury in concussed female student athletes than their male counterparts.⁵ Dolle and coworkers⁹ (2018) attributed this sex-specific dissociation in pathology to more robust axonal architecture in the male brain. Although other mechanisms have been proposed to explain the higher incidence of SRC and slower recovery trajectory for PCS in female athletes, the evidence for a sex difference in axonal architecture is compatible with DTI findings reported by Koerte and associates⁵ and is represented in the current study.

Limitations of this study include imaging on only a single occasion. Nevertheless, our findings underscore the 2016 Consensus Statement by the Concussion in Sport Group, which acknowledges that neurobiological recovery lags clinical recovery.¹¹ This point is especially salient in adolescents who are injured during a period of dynamic brain maturation. At the time of her first SRC, Jane was 15 years old, an age consistent with partially mature prefrontal cortex and ongoing myelination of the WM, which continues in a linear trajectory through age 30.²⁸ Despite her clinical recovery from SRC2 within 10 days, there is a possibility of remote, subtle effects on brain aging, cognitive efficiency, and motor skills, especially in view of extensive exposure to soccer.²⁹ In summary, the aberrant structural and functional connectivity belies Jane's clinical recovery from SR2.

The paucity of long-term, longitudinal imaging follow-up data for concussed adolescent athletes is a limitation for interpreting these findings. The extant literature does not provide an evidenced-based guideline concerning the time course for aberrant structural or functional connectivity after an SRC in an adolescent. We acknowledge the possibility that FA declined or failed to rise consistent with typical maturation after SRC1. It also is difficult to weigh the contribution of repetitive head impacts over this athlete's 15 years of exposure to soccer. It is plausible, however, that the cumulative effects of exposure contributed to the persistent disruption of structural and functional connectivity. This report also supports the 2016 Consensus Guidelines,¹¹ which encourage investigation of variability in recovery, especially individual differences and studying sex-specific variables related to concussion.

This dissociation between aberrant structural and functional brain connectivity, despite clinical recovery from SRC2 according to consensus guidelines, also raises the possibility that the clinical tools currently used to manage sports concussion may lack sensitivity to detect residual deficits a month or later after injury. For example, assessments of SRC do not typically include measurement of transcallosal reaction time even though imaging studies have implicated the vulnerability of this commissural tract.²

Conclusion

The current study shows objective imaging evidence of residual brain injury in a female adolescent athlete after her second concussion sustained while playing soccer and extensive exposure to this sport since early childhood. The full implications of our findings remain to be elucidated in long-term longitudinal studies of sports concussion in adolescent athletes. These data encourage examination of variability in brain imaging findings, especially in athletes whose recovery from concussion is prolonged.

Author Disclosure Statement

No competing financial interests exist.