Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jun 1.
Published in final edited form as: Brain Imaging Behav. 2016 Jun;10(2):559–568. doi: 10.1007/s11682-015-9407-3

Alterations in default-mode network connectivity may be influenced by cerebrovascular changes within one week of sports related concussion in college varsity athletes: a pilot study

Adam R Militana 1, Manus J Donahue 1, Allen K Sills 2, Gary S Solomon 2, Andrew J Gregory 3, Megan K Strother 1, Victoria L Morgan 1
PMCID: PMC4644725  NIHMSID: NIHMS691107  PMID: 25972119

Abstract

The goal of this pilot study is to use complementary MRI strategies to quantify and relate cerebrovascular reactivity, resting cerebral blood flow and functional connectivity alterations in the first week following sports concussion in college varsity athletes. Seven college athletes (3F/4M, age = 19.7 ± 1.2 yrs) were imaged 3–6 days following a diagnosed sports related concussion and compared to eleven healthy controls with no history of concussion (5M/6F, 18–23 yrs, 7 athletes). Cerebrovascular reactivity and functional connectivity were measured using functional MRI during a hypercapnia challenge and via resting-state regional partial correlations, respectively. Resting cerebral blood flow was quantified using arterial spin labeling MRI methods. Group comparisons were made within and between 18 regions of interest. Cerebrovascular reactivity was increased after concussion when averaged across all regions of interest (p=0.04), and within some default-mode network regions, the anterior cingulate and the right thalamus (p<0.05) independently. The FC was increased in the concussed athletes within the default-mode network including the left and right hippocampus, precuneus and ventromedial prefrontal cortex (p<0.01), with measures being linearly related to cerebrovascular reactivity in the hippocampus in the concussed athletes. Significant resting cerebral blood flow changes were not detected between the two groups. This study provides evidence for increased cerebrovascular reactivity and functional connectivity in the medial regions of the default-mode network within days of a single sports related concussion in college athletes. Our findings emphasize the utility of complementary cerebrovascular measures in the interpretation of alterations in functional connectivity following concussion.

Keywords: concussion, cerebrovascular reactivity, cerebral blood flow, functional MRI, functional connectivity

Introduction

Neuroimaging has traditionally played a limited role in the management of sports-related concussion. This is because the clinical symptoms of concussion in the acute setting likely represent functional manifestations of currently undetectable structural injury (McCrory et al. 2013). Both neurocognitive and physiologic alterations are believed to contribute to the functional impairments following concussion. While the neurocognitive alterations following concussion have been extensively studied and play a significant role in clinical management, objective physiologic measures have not been reliably identified to inform clinical decisions (Grindel et al. 2001).

Using blood oxygenation level dependent (BOLD) functional MRI (fMRI) (Logothetis et al. 2001; Ogawa et al. 1990) images acquired at rest, synchronization in low frequency blood oxygen oscillations can quantify functional connectivity (FC) across brain networks (Biswal et al. 1995; Rogers et al. 2007). This measure has been highly studied after concussion as a potential neurophysiological biomarker of injury. Studies of FC include acquisitions ranging from 10 days (Johnson et al. 2012; Zhang et al. 2012) to years (Palacios et al. 2013) after injury with several reporting general similarities in decreased FC across parts of the default-mode network (Johnson et al. 2012; Mayer et al. 2011; Y. X. Zhou et al. 2012). However, at day 1 after injury increases in some FC regions have been detected (Zhu et al. 2014). Since FC is an indirect probe of neuronal activity mediated by cerebrovascular function, the potential interaction of other cerebrovascular changes such as cerebral blood flow and vascular reactivity to the BOLD measure of FC after concussion should be considered (Lu et al. 2008; Liu et al. 2013).

Alterations in cerebrovascular physiology following concussion have long been suspected to play a role in the symptoms of concussion (Junger et al. 1997; Strebel et al. 1997; McQuire et al. 1998). Cerebral blood flow (CBF) has been found to be decreased immediately following concussion (Maugans et al. 2012) and months to years after injury in some cases (Bonne et al. 2003; Ge et al. 2009; Bartnik-Olson et al. 2014). Cerebrovascular reactivity (CVR), the change in CBF in response to a stimulus, has been less frequently studied (Becelewski and Pierzchala 2003; Len et al. 2011; Len et al. 2013). Len and colleagues have reported abnormal CVR following mild traumatic brain injury in two separate studies using MCA Doppler velocity measurements and breath-hold challenges. Most recently, they describe subject specific differences in CVR at post injury day 2 when compared with baseline and post-injury days 4 and 8 (Becelewski and Pierzchala 2003; Len et al. 2011; Len et al. 2013).

Using multi-modal MRI, measurements of FC, CBF and CVR can be obtained in the same imaging session. Arterial spin labeling (ASL) (Jiang et al. 2010; Chen et al. 2011; Wang et al. 2003; Detre et al. 1992) can be used to quantify resting CBF by comparing images where the in flowing blood is tagged with a radiofrequency pulse to invert the magnetization of blood water to images without the inversion applied. In the fMRI measurement of CVR, hypercapnia is used to elicit blood vessel dilation resulting in blood oxygenation changes detectable with BOLD fMRI (Bright and Murphy 2013; Murphy et al. 2011). Therefore, the goal of this exploratory, pilot study is to use these complementary MRI strategies to quantify and relate the CVR, CBF and FC alterations in the first week following concussion in college athletes. We hypothesize that the concurrent measures will provide a more accurate assessment and interpretation of the cerebrovascular changes early in recovery of sports concussion.

Materials and Methods

Subjects

Seven varsity college athletes were enrolled in this study within one week of experiencing a sports related concussion (3F/4M, age = 19.7 ± 1.2 yrs) (Table 1). All athletes were evaluated on the field by a certified athletic trainer for signs or symptoms of a concussion (McCrory et al. 2013). Athletes with positive findings were referred to a sports medicine physician who confirmed the diagnosis of concussion and entered them into the treatment protocol set by the Vanderbilt Sports Concussion Center. Exclusion criteria were posttraumatic amnesia > 24 hours, loss of consciousness > 30 minutes and prior history of concussion. No structural injury related to concussion was evident on conventional T1 and T2 weighted imaging reviewed by a board certified neuroradiologist. No subjects experienced persistent symptoms longer than 2 weeks.

Table 1.

Subject information

ID Age (yrs) M/F hand sport days since injury Headache score Rivermead total
Pat01 19 F R lacrosse 4 3 25
Pat02 20 M R soccer 6 0 1
Pat03 18 F R lacrosse 6 1 7
Pat04 19 M R baseball 3 2 7
Pat05 22 M R soccer 5 2 7
Pat06 20 F R soccer 6 1 16
Pat07 20 M R baseball 3 3 15
Con01 21 F R volleyball n/a 0 0
Con02 18 F B volleyball n/a 0 0
Con03 20 M R none n/a 0 0
Con04 21 M R none n/a 0 0
Con05 21 F R none n/a 0 0
Con06 18 M R none n/a 0 0
Con07 20 F L basketball n/a 0 0
Con08 21 M R cross country n/a 0 0
Con09 23 M R cross country n/a 0 0
Con10 18 F R softball n/a 0 1
Con11 19 F L golf n/a 0 0
Open in a new tab

M = Male; F = Female, L = Left; R = Right; B=Bilateral handedness; Headache score (maximum possible = 4) and Rivermead total (maximum possible = 64) come from Rivermead Post-Concussion Symptoms Questionnaire (King et al. 1995).

In addition, eleven healthy college students with no history of concussion were enrolled (6F/5M, age = 20.0 ± 1.6 yrs). Seven of these controls were also varsity college athletes at the time of enrollment. All subjects gave written informed consent per Institutional Review Board guidelines. Symptoms were measured by the Rivermead Post Concussion Symptoms Questionnaire (King et al. 1995) in all subjects at the time of scanning. This assessment asks the subject to rate 16 individual symptoms over the last 24 hours from 0 (not experienced at all) to 4 (a severe problem). The rating for “Headaches” and the total over all symptoms for each subject are included in Table 1.

MRI Imaging

All imaging was performed on a 3T MRI scanner (Philips Healthcare, Best, Netherlands) using a 32-channel head coil. The protocol for each subject included three structural acquisitions: 1) three-dimensional, T1 weighted high-resolution whole brain image for inter-subject normalization (1 × 1 × 1 mm3) and identification of structural abnormalities, 2) T2 weighted turbo spin echo, high resolution axial full brain image (0.5 × 0.5 × 2.5 mm3) for identification of structural abnormalities, and 3) two-dimensional, T1 weighted high-resolution axial full brain image (1 × 1 × 4 mm3) in the same slice locations as the fMRI for functional to 3D data registration. Physiological monitoring of cardiac and respiratory fluctuations was performed at 500 Hz using the MRI scanner integrated pulse oximeter and the respiratory belt.

For the measurement of FC, a T2* weighted gradient echo, echo-planar fMRI image series was acquired at rest with eyes closed – matrix 80 × 80, FOV = 240 mm, 34 axial slices, TE = 35 ms, TR = 2 sec, slice thickness = 3.5 mm with 0.5 mm gap, 300 volumes (10 minutes). The same T2* weighted sequence was used to acquire images with a hypercapnia challenge for the measurement of CVR, with the only change being in length of scan (270 volumes). The 9 minute hypercapnia challenge consisted of the subject alternating breathing medical air (13 L/min) and medical air with 5% CO2 (5% CO2/21% O2/74% N2, 13 L/min) in a pattern of 1-2-2-2-2 minute blocks each through a face mask covering the nose and mouth. End-tidal CO2 was monitored via a nasal cannula and recorded at approximately 20 second intervals. The average change in expired CO2 (mmHg) between the blocks of medical air and the CO2 challenge was computed across the entire scan.

For the quantification of resting cerebral blood flow (CBF), a pseudo continuous arterial spin labeling (pCASL) (Chen et al. 2011; Dai et al. 2008) approach was utilized (matrix 80 × 80, FOV = 240 mm × 240 mm, 17 axial slices, 7 mm slice thickness with no gap, echo time = 13.78 ms, SENSE =2.5 ). A total of 35 pairs of interleaved labeled and control images were acquired. An M0 image to measure equilibrium magnetization was acquired in the same slice locations with a TR = 20 sec, and echo time = 11.89 msec.

The order of the scans remained the same for all subjects. The session started with the subject wearing the face mask and cannula. The scan order was the following: 1) survey, 2) two-dimensional T1 weighted scan with orientation matching the CVR fMRI, and 3) CVR fMRI scan. The mask and cannula were then removed from the subject. The scanning resumed with 4) survey 5) three-dimensional T1 weighted scan, 6) two-dimensional T1 weighted scan with orientation matching the FC fMRI scan, 7) FC fMRI scan, 8) CBF scan, 9) M0 scan, and 10) T2 weighted scan. The session concluded with other scans not included in these analyses.

Cerebrovascular Reactivity and Functional Connectivity Analysis

The CVR and the FC fMRI images were corrected for slice timing effects and motion occurring between the fMRI scans using SPM8 software (http://www.fil.ion.ucl.ac.uk/spm/software/spm8/). The maximum translation in the x, y, and z directions and the maximum rotation about the x, y, and z axes were computed for each subject. Then images were corrected for physiological noise using a RETROICOR protocol (Glover et al. 2000) using the measured cardiac and respiratory time series. Images were then spatially normalized to the Montreal Neurological Institute (MNI) template and smoothed with a 6 × 6 × 6 mm3 full width, half maximum kernel. The T1 weighted, three-dimensional, high-resolution MRI image was segmented into its gray matter, white matter and cerebrospinal fluid components using SPM8.

The FC images were temporally low pass filtered at 0.1 Hz (Cordes et al. 2001). Then the average time series within each region of interest described below was calculated. FC was computed between all pairs of regions using the partial correlation between the average time series in two regions, with the six motion time series and the average white matter time series as confounds.

Using the CVR images, a voxel-wise analysis was performed to determine the average percent signal change between the baseline blocks (medical air) and the stimulus blocks (medical air with 5% CO2) per voxel. The baseline average was computed from the entire first baseline minute and the last 90 seconds of each of the subsequent two baseline blocks of 2 minutes each. The stimulus signal average was computed from the last 90 seconds of each stimulus block. The first 30 seconds of each block (with the exception of the first minute) was ignored to allow the appropriate gas mixture to reach the subject and respiration to occur. The percent BOLD signal change of each voxel was computed and divided by the average expired mmHg CO2 described above. This was used as quantification of CVR (%ΔBOLD/ΔmmHg) for comparisons between groups (Fig 1).

Fig. 1.

Fig. 1

Open in a new tab

Example images and parametric maps of T1-Weighted (T1W), T2-Weighted (T2W), resting cerebral blood flow (CBF) and cerebrovascular reactivity (CVR) in individual subjects Con05 and Pat02. CBF is measured in ml/100g tissue/min. CVR is reported in % signal change/end tidal CO2 (mmHg).

Resting Cerebral Blood Flow Analysis

The 35 pairs of pCASL labeled and control images were first realigned using SPM8 software. Then a pairwise subtraction was used to determine perfusion difference images. The voxel-wise CBF (ml/100g tissue/minute) was quantified using the single blood compartment model (Chen et al. 2011; Dai et al. 2008; Buxton et al. 1998):

f=λ∗ΔM2α∗M0∗T1b∗(e-ωT1b-e-(τ+ω)/T1b)) (1)

where λ = 0.98 ml/g blood to water partition coefficient, ΔM = the mean of the perfusion difference images, α = 0.85 assumed label efficiency, τ = label duration of 1.65 sec, T1b = 1.7 sec T1 of blood water at 3T. The ω = the post labeling delay of 1.65 sec corrected for the individual slice acquisition times, and M0 was taken as the maximum signal in the cerebral spinal fluid in the M0 image divided by the water density (0.87). The resulting voxel-wise map of CBF was spatially normalized to the MNI template and resampled to 3 × 3 × 3 mm3 to match the resolution of the CVR maps (Fig 1). Gray matter maps of CBF were determined by masking this map with the individual’s gray matter segmented T1 image normalized to MNI space.

Regions of Interest

Eighteen regions of interest were defined across the brain in all subjects for comparison of the cerebrovascular measures. Sixteen of these were chosen from three well-defined networks studied previously with functional MRI – default-mode network (DMN) (Buckner et al. 2008), dorsal attention network (DAN) (Gao and Lin 2012), and frontal parietal control network (FPC) (Gao and Lin 2012). Previous studies have reported changes in functional connectivity in the DMN in mild traumatic brain injury (Johnson et al. 2012; Zhang et al. 2012; Sharp et al. 2011). It is believed that the DAN is anti-correlated with the DMN, and that the balance between these two functional networks is controlled by the FPC (Gao and Lin 2012; Sridharan et al. 2008). By defining widespread regions based on these networks, cerebrovascular characteristics throughout the brain can be investigated, and possibly interpreted in the context of this balance between networks. The MNI coordinates for each region of interest in the three networks were taken from Table 1 in (Gao and Lin 2012) and identified using the aal atlas in WFU PickAtlas software (Maldjian et al. 2003; Tzourio-Mazoyer et al. 2002). The DMN regions included: hippocampus left, hippocampus right, precuneus (left and right combined), angular left (left posterior inferior parietal lobe), angular right (right posterior inferior parietal lobe), frontal middle orbital (left and right combined as ventromedial prefrontal cortex), and frontal superior medial (left and right combined as dorsomedial prefrontal cortex). The FPC regions included: cingulum anterior (left and right combined), parietal inferior left, parietal inferior right, frontal middle left (dorsolateral prefrontal cortex left), frontal middle right (dorsolateral prefrontal cortex right). The DAN regions included: occipital middle left (middle temporal area left), occipital middle right (middle temporal area right), parietal superior left (intra-parietal sulcus left), parietal superior right (intra-parietal sulcus right). The left and right thalamus were also defined. These were chosen based on findings of disruption of thalamic and thalamic-cortical functional networks in mTBI (Tang et al. 2011; Y. Zhou et al. 2013). Average CBF and CVR in each region of interest were computed.

Statistical Comparisons

The linear correlation between the concussed athletes’ Rivermead headache and total scores and days after injury were computed using a Pearson’s correlation. A two sample t-test was used to compare the CVR and the CBF values in each region between the concussion group (n=7) and the control group (n=11), and the subset of the control subjects who were athletes (n=7). Where significant differences were found, the effect size was computed as the Cohen’s d statistic (Cohen 1988). FC was compared between the control group and the concussion group for each pair of regions using a two sample t-test. The Cohen’s d statistic was again computed for any significant difference between groups. Any spatially overlapping findings of significant changes between groups in CVR, CBF and FC were investigated with a Pearson’s correlation. This was tested in the control group and the concussion group separately.

Results

Rivermead headache score reported in the concussed athletes was negatively linearly correlated with days after injury (p=0.02), while total Rivermead scores were not. One concussed athlete (pat04) was given 95% O2 with 5% CO2 instead of the medical air with 5% CO2, so was not analyzed for the CVR investigation due to the potential effect of the increased oxygen on the BOLD fMRI signal (Hare et al. 2013), resulting in six patients for the CVR analysis. The CVR averaged across all of the regions of interest was increased in the concussed athletes compared to the healthy controls (p=0.04), and compared to the subset of healthy control athletes (p=0.03). When looking at individual regions, the statistically significant increases between groups were most concentrated in the DMN and the anterior cingulate in the FPC (Fig 2). The CVR averaged over all of the DMN regions was increased in the concussed athletes compared to the healthy controls (p=0.04), and compared to the subset of healthy control athletes (p=0.02). Additionally, regions which did not reach statistical significance demonstrated an average increase in CVR in the concussed athletes compared to all controls or athlete controls (Fig 3A). The Cohen’s d effect size for each statistically significant difference is also given in Fig 2B.

Fig. 2.

Fig. 2

Open in a new tab

Regional analysis of CVR and CBF in athletes with concussion. Fig 2A contains the uncorrected p-values of group differences. Yellow indicates statistically significant difference at p<0.05 (uncorrected). Fig 2B contains the Cohen’s d statistic (Cohen 1988) of those comparisons with statistically significant differences. CVR = cerebrovascular reactivity, CBF = resting cerebral blood flow, pat = patients with concussion, all = all controls including athletes and non-athletes, ath = athlete controls only, L = left, R = right, POS = posterior, INF = inferior, DMN = regions in default-mode network, FPC = regions in frontal-parietal control network, DAN = regions in dorsal attention network, GM = all gray matter. Numbers to the left of each ROI correspond to number in x-axes in Fig 3 and in Fig 4.

Fig. 3.

Fig. 3

Open in a new tab

Cerebrovascular reactivity (CVR) (A) and resting cerebral blood flow (CBF) (B) in sports concussion. The x-axis represents each region of interest with the numbers corresponding to the numbers in Fig 2. The y-axis is the average across all subjects in the group in the region of interest. DMN = regions in default-mode network, FPC = regions in frontal-parietal control network, DAN = regions in dorsal attention network.

The CBF averaged across all of the regions of interest was not changed in the concussed athletes compared to the healthy controls (p=0.51), nor compared to the healthy control athletes (p=0.11). Fig 3B shows that the average CBF in each region in the concussed athletes tended to be lower than all controls, but did not reach statistical significance. Similarly, when compared to the athlete controls, the concussed athletes tended to have an average decrease in CBF. However, these decreases were statistically significant (p<0.05) only in the left anterior inferior parietal lobe of the FPC and the left middle temporal area of the DAN in the concussed athletes when compared to the healthy control athletes only (Fig 2).

The FC was increased in the concussed athletes compared to the healthy controls across three pairs of regions in the DMN including the left and right hippocampus, precuneus and ventromedial prefrontal cortex (p<0.01) (Fig 4A, solid lines. Two other pairs including the DMN and the anterior cingulate in the FPC were also increased in the concussed athletes (p<0.05) (Fig 4A, dashed lines). Compared to the subset of healthy athletes only, the concussed athletes had increased FC in the DMN between the right hippocampus and the ventromedial prefrontal cortex (p<0.01). Averaged across all DMN pairs of regions there was a trend towards increase FC in concussed athletes compared to healthy controls (p=0.06).

Fig. 4.

Fig. 4

Open in a new tab

Functional connectivity changes in sports concussion. A) FC increases in concussed athletes compared to healthy controls. Nodes of different networks are indicated by different colors as described in the legend. Nodes of increased CVR in concussed athletes are filled in yellow. Numbers of nodes correspond to the numbers in Fig 2. B) Linear relationship between FC in right hippocampus to precuneus and CVR in right hippocampus in controls (p=0.03) and concussed athletes (p=0.03). C) Linear relationship between FC in left hippocampus to precuneus and CVR in left hippocampus in controls (p>0.05) and concussed athletes (p=0.001). Athlete and non-athlete controls are shown in different colors, but regression performed across group as a whole.

The CVR in the hippocampus was linearly related to the FC between the hippocampus and precuneus across subjects, but this relationship was different between the healthy controls and the concussed athletes. In the right hippocampus, the healthy controls showed a linear decrease in FC between right hippocampus and precuneus with increase in right hippocampus CVR (Fig 4B) (p=0.03, d=1.7), while the concussed athletes had an increase in FC with increase CVR (Fig 4B) (p=0.03, d>2.0). In the left hippocampus, there was an increase in the FC between the left hippocampus and precuneus with an increase CVR in the left hippocampus in concussed athletes (Fig 4C) (p=0.001 d>2.0), but not healthy controls.

Discussion

In this study, we had three main findings. First, we observed increased CVR in response to a hypercapnia challenge in college athletes in the days following a sports-related concussion. Several regions concentrated within the DMN, the anterior cingulate and the right thalamus reached statistical significance independently. While the overall trend of our data suggests that the increase in CVR following concussion is likely a global process, the medial DMN may be particularly sensitive to concussive forces. Second, we found increases in FC between the hippocampus, precuneus and ventromedial prefrontal cortex in the DMN in concussed athletes. There was also a trend towards increase FC across the DMN as a whole in the concussed athletes. Third, we found that the increase in CVR in the hippocampus is positively correlated with the increase in FC between the hippocampus and precuneus in the concussed athletes, but not in healthy controls.

The finding of increased CVR is unusual given that CVR is usually decreased in the majority of pathologic conditions in which cerebrovascular physiology is altered with the exception of migraine headaches (Thomsen et al. 1995). This exception is particularly interesting given that headache is a commonly reported symptom in patients following concussion. We suspect that the mechanism underlying the increases in CVR we observed may be nitric oxide mediated. Upregulation of inducible nitric oxide synthase following traumatic brain injury is well established, and a study in concussed rats following mild traumatic brain injury has shown similar increases in nitric oxide synthase production (Gahm et al. 2000; Petrov et al. 2000; Wada et al. 1998; Tavazzi et al. 2007; Clark et al. 1996). Interestingly, Peebles et al. found increased levels of nitric oxide in cerebral venous samples following hypercapnia indicating that the vasodilatory response to hypercapnia challenge is nitric oxide mediated (Peebles et al. 2008). It remains to be established whether a persistently abnormal CVR has the potential to result in symptom recurrence upon return to physical activity as the partial pressure of carbon dioxide in blood is just one of many factors both systemic and local that influence cerebral blood flow during exercise (Ogoh and Ainslie 2009).

To further investigate the potential link between CVR and headache in our sample, we performed a partial correlation between Rivermead headache score and CVR across the patients in the seven regions shown to be significantly increased in patients vs. all controls in Fig 2. We used the days after injury as a confound. Of those regions, the right hippocampus region of the DMN (p=0.005), the right thalamus (p=0.04), average of ALL regions (p=0.03), and the average across the DMN (p=0.03) showed significant negative correlation with headache after correcting for days after injury. Considering the seven regions tested, only the right hippocampus exceeds the p<0.05 threshold corrected for multiple comparisons (0.05/7=0.007).

While we detected a trend of decreased CBF across most regions of the brain (Fig 3B), we found no statistically significant difference in CBF between concussed athletes and controls. In a pediatric population, CBF was decreased less than 72 hours following concussion, but was not different than matched controls by day 14 (Maugans et al. 2012). Similar findings of focal decreased CBF were reported in college football players (Meier et al. 2015). Other reports of decreased CBF have been detected in the chronic phase after injury when symptoms persist (Bonne et al. 2003; Ge et al. 2009; Bartnik-Olson et al. 2014). Our inability to detect significant decreases in CBF in the concussed athletes may be due in part to our control group. Our data shows a strong trend of increased CBF in athlete controls over non-athlete controls. Also, a few regions showed decreases when comparing concussed athletes to athlete controls only. These suggest that a larger control sample consisting only of athletes may be required to detect these differences (as in (Meier et al. 2015)). In addition, the athletes in both the Maugans et al. (Maugans et al. 2012) and Meier et al. (Meier et al. 2015) studies included concussed football players in which repeated concussive or subconcussive impacts may play a role.

We report a significant and localized finding of increased hippocampal, precuneus and ventromedial prefrontal FC in the concussed athletes over the healthy controls. It is difficult to relate these findings to previous literature. Early studies performed at approximately 30 days post injury (Mayer et al. 2011), 10 days post injury (Johnson et al. 2012), and two months post injury (Y. X. Zhou et al. 2012) detected decreases in resting FC in some DMN regions. However, a longitudinal study found increases in FC compared to controls in the DMN at day 1 followed by significant decreases at day 10 (Zhu et al. 2014), indicating a progression of changing FC in the DMN over the first few weeks after injury. Our findings of increased FC in parts of the DMN in days 3 to 6 may suggest that the decrease in FC occurs after the first week. Our findings are also consistent with those detected immediately following potential subconcussive injury. Johnson et al.(Johnson et al. 2014) looked at differences between pre-game and post-game MRI measures of FC and found that in those players without history of concussion, some increases in FC were detected between the supramarginal gyrus, the orbitofrontal cortex, the retrospenial cortex and the medial prefrontal cortex similar to our findings. Another study found increased DMN connections of FC in football players imaged at multiple sessions throughout a single season (Abbas et al. 2014), presumably within a day after subconcussive impacts. While our data reflects early concussion injury, it is possible that recent (within 24–48 hours) subconcussive impacts may result in similar pathophysiology.

The spatially overlapping, linear relationship of FC between the hippocampus and precuneus and CVR in the hippocampus has not been previously identified. Generally, decreases (not increases) in FC are attributed to network impairment, however, these findings suggest that the increase in FC we detect may be partially related to a pathologic hyper-reactivity of the vasculature across the DMN in the concussed athletes. One possible mechanism for this relationship is that that the increased CVR results in an increase in the power of the resting BOLD oscillations at the same frequency in two regions. This may decrease the contribution of other discrepant frequency oscillations in the FC between the two regions resulting in an overall increase in FC. While the validation of this mechanism is beyond the scope of this study, this relationship underscores the importance of interpretation of functional connectivity results following concussion in the context of concurrent cerebrovascular measurements. This may be particularly important in interpreting progression of FC alterations as previously discussed because longitudinal measures of CVR and CBF may also be required. Task-based fMRI studies where percent BOLD changes in the hippocampus are increased after concussion may be similarly influenced if underlying cerebrovascular pathophysiology is not considered (Slobounov et al. 2010).

The primary limitation of this study is the small sample size. Therefore, we also report the Cohen’s d effect sizes of our findings. Our analyses of CVR and FC yielded d values greater than 1.0, and in many cases greater than 1.5. At d=1.0 and 1.5, the mean of the concussed athletes group is expected to be at the 84% and 93.3% of the control group, respectively, and is considered a large effect size (Cohen 1988). Similarly, it must be noted that due to the exploratory nature of this work with multiple scans, regions and symptoms, our statistics are not corrected for multiple comparisons unless stated. Our intention is that these findings can be used to define more specific hypotheses to be tested in future studies with larger sample sizes.

Another potential limitation is the mix of athletes and non-athletes in our control population. Therefore, we reported the results of the concussed athletes compared to the entire control population, as well as to the athlete controls subgroup. These controls are used in lieu of baseline measures (before concussion) on the concussed athletes. In addition, our time window of acquisition of 3–6 days post-injury is earlier than several studies discussed, but the variability of recovery or injury within this window is unknown. When interpreting the CVR changes, we considered that a lower baseline CBF level may increase CVR by increasing percent signal change during hypercapnia challenge without increasing peak signal. However, the small decreases in CBF do not account for the magnitude of percent BOLD fMRI signal change we observed in our measurement of CVR. Additionally, there is no overlap in individual regions in which statistically significant increases in CVR and decreased CBF were observed. Finally, we considered that the measurement of resting FC could be artifactually influenced by the hypercapnia challenge used for the CVR measurement. We believe this possibility is accounted for as not only was the scan order the same for all patients (CVR hypercapnia challenge occurring approximately 20 minutes prior to the FC fMRI scan), the BOLD signal returned to baseline following the hypercapnia challenge in all subjects. Furthermore, the facemask was removed following the hypercapnia challenge such that there was no possibility of breathing unaccounted for residual gas within the tube in the interim between the two scans.

In conclusion, this exploratory, pilot study provides preliminary evidence for increased CVR in response to a hypercapnia challenge within days of a single sports related concussion in college athletes, without significant changes in resting CBF. These increases in CVR are accompanied by linearly related increases in FC in medial DMN regions, which suggests that pathologic cerebrovascular hyper-reactivity may play a role in increasing the FC in these regions. Our findings suggest that complementary cerebrovascular measures should be considered in the interpretation of alterations in functional connectivity following concussion.

Acknowledgments

This work was supported in part by Vanderbilt CTSA grant UL1 TR000445 from NCRR/NIH (Morgan).

Funding: This work was supported in part by Vanderbilt CTSA grant UL1 TR000445 from NCRR/NIH (Morgan).

Footnotes

Conflict of Interest: The authors declare that they have no conflict of interest.

Compliance with Ethical Standards

Ethics approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the Vanderbilt University and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all participants included in this study.

References

  1. Abbas K, Shenk TE, Poole VN, Breedlove EL, Leverenz LJ, Nauman EA, et al. Alteration of Default Mode Network in High School Football Athletes Due to Repetitive Subconcussive Mild Traumatic Brain Injury: A Resting-State Functional Magnetic Resonance Imaging Study. Brain Connect. 2014 doi: 10.1089/brain.2014.0279. [DOI] [PubMed] [Google Scholar]
  2. Bartnik-Olson BL, Holshouser B, Wang H, Grube M, Tong K, Wong V, et al. Impaired Neurovascular Unit Function Contributes to Persistent Symptoms after Concussion: A Pilot Study. [Article] Journal of Neurotrauma. 2014;31(17):1497–1506. doi: 10.1089/neu.2013.3213. [DOI] [PubMed] [Google Scholar]
  3. Becelewski J, Pierzchala K. Cerebrovascular reactivity in patients with mild head injury. Neurol Neurochir Pol. 2003;37(2):339–350. [PubMed] [Google Scholar]
  4. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn Reson Med. 1995;34(4):537–541. doi: 10.1002/mrm.1910340409. [DOI] [PubMed] [Google Scholar]
  5. Bonne O, Gilboa A, Louzoun Y, Kempf-Sherf O, Katz M, Fishman Y, et al. Cerebral blood flow in chronic symptomatic mild traumatic brain injury. [Article] Psychiatry Research-Neuroimaging. 2003;124(3):141–152. doi: 10.1016/s0925-4927(03)00109-4. [DOI] [PubMed] [Google Scholar]
  6. Bright MG, Murphy K. Reliable quantification of BOLD fMRI cerebrovascular reactivity despite poor breath-hold performance. [Article] Neuroimage. 2013;83:559–568. doi: 10.1016/j.neuroimage.2013.07.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Buckner RL, Andrews-Hanna JR, Schacter DL. The brain’s default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci. 2008;1124:1–38. doi: 10.1196/annals.1440.011. [DOI] [PubMed] [Google Scholar]
  8. Buxton RB, Frank LR, Wong EC, Siewert B, Warach S, Edelman RR. A general kinetic model for quantitative perfusion imaging with arterial spin labeling. [Article] Magnetic Resonance in Medicine. 1998;40(3):383–396. doi: 10.1002/mrm.1910400308. [DOI] [PubMed] [Google Scholar]
  9. Chen YF, Wang DJJ, Detre JA. Test-Retest Reliability of Arterial Spin Labeling With Common Labeling Strategies. [Article] Journal of Magnetic Resonance Imaging. 2011;33(4):940–949. doi: 10.1002/jmri.22345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Clark RSB, Kochanek PM, Schwarz MA, Schiding JK, Turner DS, Chen MZ, et al. Inducible nitric oxide synthase expression in cerebrovascular smooth muscle and neutrophils after traumatic brain injury in immature rats. [Article] Pediatric Research. 1996;39(5):784–790. doi: 10.1203/00006450-199605000-00007. [DOI] [PubMed] [Google Scholar]
  11. Cohen J. Statistical power analysis for the behavioral sciences. 2. Hillsdale, NL: Lawrence Earlbaum Associates; 1988. [Google Scholar]
  12. Cordes D, Haughton VM, Arfanakis K, Carew JD, Turski PA, Moritz CH, et al. Frequencies contributing to functional connectivity in the cerebral cortex in “resting-state” data. [Article] American Journal of Neuroradiology. 2001;22(7):1326–1333. [PMC free article] [PubMed] [Google Scholar]
  13. Dai WY, Garcia D, de Bazelaire C, Alsop DC. Continuous Flow-Driven Inversion for Arterial Spin Labeling Using Pulsed Radio Frequency and Gradient Fields. [Article] Magnetic Resonance in Medicine. 2008;60(6):1488–1497. doi: 10.1002/mrm.21790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Detre JA, Leigh JS, Williams DS, Koretsky AP. PERFUSION IMAGING. [Article] Magnetic Resonance in Medicine. 1992;23(1):37–45. doi: 10.1002/mrm.1910230106. [DOI] [PubMed] [Google Scholar]
  15. Gahm C, Holmin S, Mathiesen T. Temporal profiles and cellular sources of three nitric oxide synthase isoforms in the brain after experimental contusion. [Article] Neurosurgery. 2000;46(1):169–177. [PubMed] [Google Scholar]
  16. Gao W, Lin WL. Frontal parietal control network regulates the anti-correlated default and dorsal attention networks. [Article] Human Brain Mapping. 2012;33(1):192–202. doi: 10.1002/hbm.21204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ge YL, Patel MB, Chen Q, Grossman EJ, Zhang K, Miles L, et al. Assessment of thalamic perfusion in patients with mild traumatic brain injury by true FISP arterial spin labelling MR imaging at 3T. [Article] Brain Injury. 2009;23(7–8):666–674. doi: 10.1080/02699050903014899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Glover GH, Li TQ, Ress D. Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn Reson Med. 2000;44(1):162–167. doi: 10.1002/1522-2594(200007)44:1<162::aid-mrm23>3.0.co;2-e. [DOI] [PubMed] [Google Scholar]
  19. Grindel SH, Lovell MR, Collins MW. The assessment of sport-related concussion: The evidence behind neuropsychological testing and management. [Article] Clinical Journal of Sport Medicine. 2001;11(3):134–143. doi: 10.1097/00042752-200107000-00003. [DOI] [PubMed] [Google Scholar]
  20. Hare HV, Germuska M, Kelly ME, Bulte DP. Comparison of CO2 in air versus carbogen for the measurement of cerebrovascular reactivity with magnetic resonance imaging. [Article] Journal of Cerebral Blood Flow and Metabolism. 2013;33(11):1799–1805. doi: 10.1038/jcbfm.2013.131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jiang L, Kim M, Chodkowski B, Donahue MJ, Pekar JJ, Van Zijl PCM, et al. Reliability and reproducibility of perfusion MRI in cognitively normal subjects. [Article] Magnetic Resonance Imaging. 2010;28(9):1283–1289. doi: 10.1016/j.mri.2010.05.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Johnson B, Neuberger T, Gay M, Hallett M, Slobounov S. Effects of Subconcussive Head Trauma on the Default Mode Network of the Brain. [Article] Journal of Neurotrauma. 2014;31(23):1907–1913. doi: 10.1089/neu.2014.3415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Johnson B, Zhang K, Gay M, Horovitz S, Hallett M, Sebastianelli W, et al. Alteration of brain default network in subacute phase of injury in concussed individuals: Resting-state fMRI study. [Article] Neuroimage. 2012;59(1):511–518. doi: 10.1016/j.neuroimage.2011.07.081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Junger EC, Newell DW, Grant GA, Avellino AM, Ghatan S, Douville CM, et al. Cerebral autoregulation following minor head injury. [Article] Journal of Neurosurgery. 1997;86(3):425–432. doi: 10.3171/jns.1997.86.3.0425. [DOI] [PubMed] [Google Scholar]
  25. King NS, Crawford S, Wenden FJ, Moss NEG, Wade DT. The Rivermead Post Concussion Symptoms Questionnaire: a measure of symptoms commonly experienced after head injury and its reliability. Journal of Neurology. 1995;242(9):587–592. doi: 10.1007/BF00868811. [DOI] [PubMed] [Google Scholar]
  26. Len TK, Neary JP, Asmundson GJ, Goodman DG, Bjornson B, Bhambhani YN. Cerebrovascular reactivity impairment after sport-induced concussion. Med Sci Sports Exerc. 2011;43(12):2241–2248. doi: 10.1249/MSS.0b013e3182249539. [DOI] [PubMed] [Google Scholar]
  27. Len TK, Neary JP, Asmundson GJG, Candow DG, Goodman DG, Bjornson B, et al. Serial monitoring of CO2 reactivity following sport concussion using hypocapnia and hypercapnia. [Article] Brain Injury. 2013;27(3):346–353. doi: 10.3109/02699052.2012.743185. [DOI] [PubMed] [Google Scholar]
  28. Liu PY, Hebrank AC, Rodrigue KM, Kennedy KM, Section J, Park DC, et al. Age-related differences in memory-encoding fMRI responses after accounting for decline in vascular reactivity. [Article] Neuroimage. 2013;78:415–425. doi: 10.1016/j.neuroimage.2013.04.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A. Neurophysiological investigation of the basis of the fMRI signal. [Article] Nature. 2001;412(6843):150–157. doi: 10.1038/35084005. [DOI] [PubMed] [Google Scholar]
  30. Lu HZ, Zhao CG, Ge YL, Lewis-Amezcua K. Baseline blood oxygenation modulates response amplitude: Physiologic basis for intersubject variations in functional MRI signals. [Article] Magnetic Resonance in Medicine. 2008;60(2):364–372. doi: 10.1002/mrm.21686. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH. An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage. 2003;19(3):1233–1239. doi: 10.1016/s1053-8119(03)00169-1. [DOI] [PubMed] [Google Scholar]
  32. Maugans TA, Farley C, Altaye M, Leach J, Cecil KM. Pediatric Sports-Related Concussion Produces Cerebral Blood Flow Alterations. [Article] Pediatrics. 2012;129(1):28–37. doi: 10.1542/peds.2011-2083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mayer AR, Mannell MV, Ling J, Gasparovic C, Yeo RA. Functional connectivity in mild traumatic brain injury. Human Brain Mapping. 2011;32(11):1825–1835. doi: 10.1002/hbm.21151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. McCrory P, Meeuwisse WH, Aubry M, Cantu B, Dvořák J, Echemendia RJ, et al. Consensus statement on concussion in sport: the 4th International Conference on Concussion in Sport held in Zurich, November 2012. British Journal of Sports Medicine. 2013;47(5):250–258. doi: 10.1136/bjsports-2013-092313. [DOI] [PubMed] [Google Scholar]
  35. McQuire JC, Sutcliffe JC, Coats TJ. Early changes in middle cerebral artery blood flow velocity after head injury. [Article] Journal of Neurosurgery. 1998;89(4):526–532. doi: 10.3171/jns.1998.89.4.0526. [DOI] [PubMed] [Google Scholar]
  36. Meier TB, Bellgowan PF, Singh R, Kuplicki R, Polanski DW, Mayer AR. REcovery of cerebral blood flow following sports-related concussion. JAMA Neurol. 2015 doi: 10.1001/jamaneurol.2014.4778. [DOI] [PubMed] [Google Scholar]
  37. Murphy K, Harris AD, Wise RG. Robustly measuring vascular reactivity differences with breath-hold: Normalising stimulus-evoked and resting state BOLD fMRI data. [Article] Neuroimage. 2011;54(1):369–379. doi: 10.1016/j.neuroimage.2010.07.059. [DOI] [PubMed] [Google Scholar]
  38. Ogawa S, Lee TM, Nayak AS, Glynn P. Oxygenation-sensitive contrast in magnetic-resonance image of rodent brain at high magnetic-fields [Article] Magnetic Resonance in Medicine. 1990;14(1):68–78. doi: 10.1002/mrm.1910140108. [DOI] [PubMed] [Google Scholar]
  39. Ogoh S, Ainslie PN. Cerebral blood flow during exercise: mechanisms of regulation. [Review] Journal of Applied Physiology. 2009;107(5):1370–1380. doi: 10.1152/japplphysiol.00573.2009. [DOI] [PubMed] [Google Scholar]
  40. Palacios EM, Sala-Llonch R, Junque C, Roig T, Tormos JM, Bargallo N, et al. Resting-state functional magnetic resonance imaging activity and connectivity and cognitive outcome in traumatic brain injury. JAMA Neurol. 2013;70(7):845–851. doi: 10.1001/jamaneurol.2013.38. [DOI] [PubMed] [Google Scholar]
  41. Peebles KC, Richards AM, Celi L, McGrattan K, Murrell CJ, Ainslie PN. Human cerebral arteriovenous vasoactive exchange during alterations in arterial blood gases. [Article] Journal of Applied Physiology. 2008;105(4):1060–1068. doi: 10.1152/japplphysiol.90613.2008. [DOI] [PubMed] [Google Scholar]
  42. Petrov T, Page AB, Owen CR, Rafols JA. Expression of the inducible nitric oxide synthase in distinct cellular types after traumatic brain injury: an in situ hybridization anal immunocytochemical study. [Article] Acta Neuropathologica. 2000;100(2):196–204. doi: 10.1007/s004019900167. [DOI] [PubMed] [Google Scholar]
  43. Rogers BP, Morgan VL, Newton AT, Gore JC. Assessing functional connectivity in the human brain by fMRI. Magn Reson Imaging. 2007;25(10):1347–1357. doi: 10.1016/j.mri.2007.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Sharp DJ, Beckmann CF, Greenwood R, Kinnunen KM, Bonnelle V, De Boissezon X, et al. Default mode network functional and structural connectivity after traumatic brain injury. [Article] Brain. 2011;134:2233–2247. doi: 10.1093/brain/awr175. [DOI] [PubMed] [Google Scholar]
  45. Slobounov SM, Zhang K, Pennell D, Ray W, Johnson B, Sebastianelli W. Functional abnormalities in normally appearing athletes following mild traumatic brain injury: a functional MRI study. [Article] Experimental Brain Research. 2010;202(2):341–354. doi: 10.1007/s00221-009-2141-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sridharan D, Levitin DJ, Menon V. A critical role for the right fronto-insular cortex in switching between central-executive and default-mode networks. [Article] Proceedings of the National Academy of Sciences of the United States of America. 2008;105(34):12569–12574. doi: 10.1073/pnas.0800005105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Strebel S, Lam AM, Matta BF, Newell DW. Impaired cerebral autoregulation after mild brain injury. [Article] Surgical Neurology. 1997;47(2):128–131. doi: 10.1016/s0090-3019(96)00459-4. [DOI] [PubMed] [Google Scholar]
  48. Tang L, Ge YL, Sodickson DK, Miles L, Zhou YX, Reaume J, et al. Thalamic Resting-State Functional Networks: Disruption in Patients with Mild Traumatic Brain Injury. [Article] Radiology. 2011;260(3):831–840. doi: 10.1148/radiol.11110014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Tavazzi B, Vagnozzi R, Signoretti S, Amorini AM, Belli A, Cimatti M, et al. Temporal window of metabolic brain vulnerability to concussions: Oxidative and nitrosative stresses - Part II. [Article] Neurosurgery. 2007;61(2):390–395. doi: 10.1227/01.neu.0000255525.34956.3f. [DOI] [PubMed] [Google Scholar]
  50. Thomsen LL, Iversen HK, Olesen J. Increased cerebrovascular PCO(2) reactivity in migraine with aura - A transcranial Doppler study during hyperventilation. [Article] Cephalalgia. 1995;15(3):211–215. doi: 10.1046/j.1468-2982.1995.015003211.x. [DOI] [PubMed] [Google Scholar]
  51. Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F, Etard O, Delcroix N, et al. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. [Article] Neuroimage. 2002;15(1):273–289. doi: 10.1006/nimg.2001.0978. [DOI] [PubMed] [Google Scholar]
  52. Wada K, Chatzipanteli K, Kraydieh S, Busto R, Dietrich WD. Inducible nitric oxide synthase expression after traumatic brain injury and neuroprotection with aminoguanidine treatment in rats. [Article] Neurosurgery. 1998;43(6):1427–1436. doi: 10.1097/00006123-199812000-00096. [DOI] [PubMed] [Google Scholar]
  53. Wang JJ, Alsop DC, Song HK, Maldjian JA, Tang K, Salvucci AE, et al. Arterial transit time imaging with flow encoding arterial spin tagging (FEAST). [Article] Magnetic Resonance in Medicine. 2003;50(3):599–607. doi: 10.1002/mrm.10559. [DOI] [PubMed] [Google Scholar]
  54. Zhang K, Johnson B, Gay M, Horovitz SG, Hallett M, Sebastianelli W, et al. Default Mode Network in Concussed Individuals in Response to the YMCA Physical Stress Test. [Article] Journal of Neurotrauma. 2012;29(5):756–765. doi: 10.1089/neu.2011.2125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhou Y, Lui YW, Zuo X-N, Milham MP, Reaume J, Grossman RI, et al. Characterization of thalamo-cortical association using amplitude and connectivity of functional MRI in mild traumatic brain injury. Journal of Magnetic Resonance Imaging. 2013 doi: 10.1002/jmri.24310. n/a-n/a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhou YX, Milham MP, Lui YW, Miles L, Reaume J, Sodickson DK, et al. Default-Mode Network Disruption in Mild Traumatic Brain Injury. [Article] Radiology. 2012;265(3):882–892. doi: 10.1148/radiol.12120748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhu DC, Covassin T, Nogle S, Doyle S, Russell D, Pearson RL, et al. A Potential Biomarker in Sports-Related Concussion: Brain Functional Connectivity Alteration of the Default-Mode Network Measured with Longitudinal Resting-State fMRI Over Thirty Days. J Neurotrauma. 2014 doi: 10.1089/neu.2014.3413. [DOI] [PubMed] [Google Scholar]

RESOURCES