Effect of exertion on blink reflex parameters in Division I football athletes

Abstract

Blink reflex assessment has been suggested as a potential marker for sport-related concussions (SRCs) screening. However, exertion level is known to affect the blink reflex, which may affect the clinical utility of sideline screening and post-concussion monitoring. This validation study of 44 male college-aged athletes found significant differences in several blink reflex parameters between baseline and active play. In assessing the differences between mild and moderate/higher intensity, as a function of heart rate, the number of oscillations was the only significantly different parameter between the two groups. When combining the two groups (mild and moderate/higher intensity) and comparing to baseline values of the blink reflex, there were significant differences in latency, initial velocity, time to open, and lid excursions. Using baseline data, the EyeStat provides a greater understanding of its application during baseline, and exercise monitoring.

Keywords:

• blink reflex
• active play
• college athletes
• oculomotor function

1. Introduction

The definition of a sport-related concussion (SRC) was recently updated during the 6th International Conference on Concussion in Sport and is defined as a “brain injury caused by a direct blow to the head or neck area” during sport-related activites, resulting in a cascade of events with acute symptoms or symptoms occuring shortly after the concussive event (Davis et al., 2023). SRCs can have a significant impact on an individual’s health, inducing physiological and metabolic changes within the brain affecting visual, vestibular, cognitive, and emotional function (Kutcher, 2014; Giza & Hovda, 2001; McCrea et al., 2003; Scorza, 2019). Emergency room survellience data from 2010-2016 cited an estimated 283,000 children sought treatment for sports-related injuries, with 2017 CDC data citing an estimated 2.5 million self-resported concussions by athletes (DePadilla et al., 2018; Sarmiento et al., 2019). Within National Collegiate Athletic Association (NCAA) sporting events, it is estimated that there are 4.13-4.47 SRCs per 10,000 athlete exposures, including practice and competitions (Chandran et al., 2022; Zuckerman et al., 2015). The typical recovery time from athlete-related concussions ranges from 7 to 14 days, with recent data from the Concussion Assessment, Research, and Education (CARE) consortium citing an average recovery time of 8.83 days (Echemendia et al., 2017; Langlois et al., 2006; McCrea et al., 2000, 2003; Wilber et al., 2021).

Concussion examinations are predominantly based on the following two parameters: (1) a physical assessment of the vestibular and ocular systems and/or a rapid neurocognitive screening such as the Standardized Assessment of Concussion (SAC) (Leddy et al., 2015) from the Sports Concussion Assessment Tool (SCAT) (Echemendia et al., 2017) as performed by a clinician, sports medicine physician, or athletic trainer; and (2) subjective reporting from the athlete commonly in the form of symptom checklists (Dubas et al., 2020; Patricios et al., 2022). Of these assessment tools utilized for concussion assessment and recovery tracking, a SRC primarily relies on the assessment of symptomology and patient reporting (Echemendia et al., 2017). However, the signs and symptoms of concussion are non-specific (Leddy et al., 2015) and may develop minutes to hours after the concussive head injury (Davis et al., 2023) thus increasing the difficulty of sideline assessments (Hyden & Petty, 2016; Leddy et al., 2015). While symptoms are critical in the diagnosis of concussion, they can be both neurogenic and psychogenic, thereby confounding the concussion diagnosis and treatment (Iverson, 2005; Pertab et al., 2018). Additionally, research identifies similiar symptoms post-exercise to those identified post-head impact, with the authors of that research citing one schould consider the effects of exercise intensity on SCAT3 severity scores (Iring-Sanchez et al., 2023). Although symptom reporting can be complicated in the concussion diagnosis, combining it with ocular assessment appears to provide improvement in the concussion diagnosis. Specifically, robustness of the SCAT3 improved when including the Vestibular/Ocular-Motor Screening (VOMS) tool with SCAT3 scoring, increasing sensitivity of acute concussion diagnosis by 9% (Ferris et al. 2021).

Evaluation of the oculomotor system is an emergent practice for the sideline and recovery assessment of concussion (Galetta et al., 2011; King et al., 2015; Yorke et al., 2017). A review of the literature cites changes in the function of eye movements as a consequence of concussion and includes impairments in the following ocular functions: saccades, smooth pursuits, optokinetic nystagmus, eye skew, and gaze control (Stuart et al., 2020). Within 24–48 hours of concussion, research cites slower smooth pursuit and faster saccadic velocity in concussed individuals as compared to controls (Murray et al., 2014), while data within 7 days of concussion showed individuals experienced lower smooth pursuit percentages and increased saccade percentages as compared to controls (Hunfalvay et al., 2021). Within 2 weeks of concussion, persons with persisting concussion symptoms demonstrated altered smooth pursuit as compared to controls (Maruta et al., 2017), and data after 3–5 months found that individuals with post-concussion syndrome (PCS) exhibited faster self-paced saccade velocity as compared to non-PCS subjects (Heitger et al., 2009). While the majority of eye-tracking studies have varied on the time of assessment from point of injury, the data provide a greater understanding on the ocular changes that occur post-concussion (Ferries et al., 2021; Stuart et al., 2020).

A novel device, referred to as the EyeStat (Blinkcns, Charleston, SC), is an FDA-approved device developed to measure changes in the blink reflex when comparing baseline and post-head impact responses. Data from the EyeStat suggest that the reflex associated with the blink response results in an involuntary response from the subject, thereby providing objective autonomic nervous system data. (Paparella et al., 2022; Tsai et al., 2017). Previous blink reflex research, as measured with the EyeStat technology, cited significant differences in latency, differential latency, log time under the threshold, number of oscillations, lid excursion, and lid velocity (see below for definitions) between baseline and post-head impact individuals (Garner et al., 2018). Each of the responses identified by the EyeStat provide a reflexive, involuntary response to a stimulus. For example, latency demonstrates the time it takes for the stimulated eyelid to blink after a stimulus, while differential latency is the contralateral eyelid response after the blinking is initiated in the stimulated eye. Thus, differences in these measured parameters from a reflexive response may provide clarity on changes in brain function as the blink reflex provides information from the cranial nerves V and VII, including connections to the pons and medulla, resulting in an extensive network within the brain (Esteban, 1999; Jerath & Kimura, 2019).

As it relates the effects of moderate to maximal physical exertion on ocular function, data cites decreased saccadic velocity after prolonged activity (Connell et al., 2017a, 2017b). This is an important consideration for any concussion assessment tool, specifically utilized on the sidelines or in recovery management that includes a physical activity protocol, as acute physical exertion affects ocular outcomes. Yet as it pertains to the blink reflex response, we were only able to find one study which assessed the effects of physical activity on blink reflex response, and this was a small cohort of a larger study focused on post-head impact effects on the reflex response (Garner et al., 2018). In this study, a cohort of healthy subjects (n = 10) demonstrated changes in the blink reflex with exercise, but the data was taken after subjects had arrived in the locker room, which was approximately 10 to 20 minutes post football practice. That study found increased latency, decreased differential latency, decreased lid excursions, faster lid velocity, decreased time to open, and decreased total blink time post exercise (Garner et al., 2018). However, considering the small sample size and the timing of the assessments, it is important to determine if a larger group of subjects exhibit differences in these same parameters immediately post-physical activity. Understanding the blink reflex response with exercise is critical within a healthy population since eye-tracking devices are becoming increasingly popular to gauge neurological health. Thus, the purpose of this study was to determine if active play, in individuals without a concussive head injury, affects blink reflex parameters on the sideline. Specifically, we examined the effects of exertion on the blink reflex in football players during a regulated spring practice scrimmage at multiple time points, during the periods when offensive and defensive players were on the sidelines between times of play.

2. Materials and methods

2.1. Study design

This prospective study was approved by the Institutional Review Board and assigned the IRB approval #1415–14. Prior to the study, participants provided consent to their enrollment in this study and were informed that they could withdraw from the study at any time. Before the study start date, all participants were required to complete a medical history questionnaire. Moreover, the participants completed an initial baseline testing session in a clinic, which included an athletic history and physical examination, balance assessment, and baseline symbol modalities tests. Along with these routine pre-season assessments, the EyeStat (Tsai et al., 2017) was used (protocol described below) to obtain baseline blink reflex data on each subject. For the active play portion of this study, the EyeStat was utilized throughout sideline play during three different pre-season scrimmages and scans were compared to the initial baselines of the subjects taken at the beginning of the season.

2.2. Participants

Healthy, non-concussed, male Division I college football players between the ages of 18 and 22 were eligible for this study. Participants were excluded if they self-reported any of the following characteristics: (1) a history of documented oculomotor abnormality; (2) a history of seizures and/or moderate to severe TBI; (3) other focal neurological deficit(s) based on history or examination; or (4) currently experiencing persistent post-concussion symptoms.

2.3. Experimental procedure

We utilized the EyeStat (Blinkcns, Charleston, SC) to monitor several blink reflex parameters (Tsai et al., 2017). Prior to use, the device was powered-on and the software was loaded. The participants then placed their faces into the facemask assembly. Upon commencement of the testing, the computer recorded up to 30 seconds of high-speed video (300 frames/sec) of the subject’s eyes while initiating blink stimuli. In order to stimulate blinking, a CO₂ cartridge delivered filtered air puffs to the right and left eyes at random intervals during the testing period. Subjects underwent repeat testing on the sidelines during three regulated spring football practice scrimmages. Testing occurred at various time points during the scrimmage when players were called off the field or between plays. More precisely, American football has offensive and defensive players and free substitution of players, with the rules of the game allowing the offensive players four plays to move the ball 10 yards forward. If they move the ball forward 10 yards, they are given four more chances to move the ball 10 yards forward and this play continues until a touchdown, field goal, or turnover occurs. If the offense is unable to move the ball 10 yards within four plays, then the defensive players take posession of the ball. Thus, when players were not involved in a specific play call and were on the sidelines, they were tested with the EyeStat device. In order to complete sideline testing, the participants sat in a chair and immediately had their heart rate (HR) assessed using an AccuMed Pulse Oximeter (AccuMed, Houston, TX). The blink reflex assessment consisted of approximately 5 brief puffs of air (randomly distributed between the right and left eye) directed near the eyes to elicit a blink reflex. Each EyeStat test lasted approximately 30 seconds, with participants being asked to undergo a maximum of 5 separate tests per session. The testing duration was 2–3 minutes from the beginning to the end of the test. While the testing itself was consistent, the time when players were tested was not similar due to the varied play calls that occurred during the scrimmage.

2.4. Main outcome measures

The proprietary data collection software evaluates and quantifies 12,000 video frames containing the eyelid movements during the 30-second testing period. The computer then runs an eyelid tracking algorithm to monitor the subject’s right and left eyelids. The eyelid tracking algorithm differentiates stimulated blinks and spontaneous blinks. The eyelid location data was outputted, along with calculations of each blink’s characteristics. The data from each scan and computed parameters were uploaded to a secure MDDS (Medical Device Data System) system for subsequent review, analysis, and reporting. For further details regarding the EyeStat technology and associated software, please review prior research (Garner et al., 2018; Tsai et al., 2017).

2.5. Parameters assessed by the EyeStat include the following

Latency: Mean elapsed time between the application of the stimulus and the detection of the blink onset in the ipsilateral eye

Differential Latency: Mean time difference between the onset of the ipsilateral and contralateral response

Delta 30: Mean time difference between when the ipsilateral and contralateral eyes moved 30 pixels below their respective tonic positions

Initial Velocity: Average eyelid speed (pixels/ms) in first 7 frames following start of eyelid movement

Time to Close: Time between the detection of the onset of the blink and the eyelid reaching a closed position

Time to Open: Time between the eyelid reaching a closed position to the eyelid returning within the open tonic position threshold

Log_e Under the Threshold: Natural log of the mean elapsed time between the eyelid reaching a closed position and the eyelid moving more than 20 pixels back towards the threshold of the tonic position

Number of Oscillations: Mean number of direction changes experienced by the eyelid between stimuli

Lid Excursions: Mean difference in pixel position between the tonic position and the location of the eyelid when it reaches a closed position

3. Statistical analysis

Descriptive univariate statistics were performed to assess the study sample at baseline. Since each participant had more than one measurement during active play, the measurement with the highest HR was selected for active play. The normality of blink reflex parameters at rest and the difference between rest and active play was assessed, Delta 30 and Time to Open were not normally distributed (Kolmogorov–Smirnov p < 0.05) hence independent samples t-test (parametric) and Mann Whitney U test (non-parametric) were used to assess groupwise differences in normally distributed and non-normally distributed variables, respectively. A paired t-test for normally distributed data and Wilcoxon Signed-rank test for non-normally distributed data was performed to see if there were any significant differences in the mean differences between rest and active play within subjects. A correlation coefficient was calculated (Pearson for normally distributed and Spearman for non-normally distributed) between each blink parameter and the HR at the active play session tosee if there were any direct correlations between HR and any of the blink reflex parameters. Effect sizes were calculated. To assess if different intensities of active play affected the blink reflex, each participant’s age appropriate maximum HR was calculated using the Karvonen equation (HR_max = 220 – age) (Peres et al., 1987) and each active play recording was categorized as mild exercise (defined as <60% of age appropriate HR_max) or moderate/higher intensity exercise (defined as ≥60% of age-appropriate HR_max) (Achten & Jeukendrug, 2003). It has been shown that HR shows an approximate linear relationship with VO₂ at submaximal intensities and can therefore be used to accurately estimate the exercise intensity. Changes in blink parameters were calculated by subtracting the subject’s baseline values from the active play values. Differences in blink parameters were compared between mild and moderate groups. A p-value of 0.05 was considered statistically significant. All statistical analyses were performed using SPSS Version 28 (IBM Corp, Armok, NY) (Stehlik-Barry & Babinec, 2017).

4. Results

Sixty-seven football players participated in this study. However, only 44 subjects had complete baseline data and at least one session of active play data. Active play data for the 44 subjects included data collected from these subjects at multiple time points during the football scrimmages and resulted in 96 total active play scans. Table 1 provides participant demographics and baseline versus active play blink parameter comparisons. When comparing the blink parameters at baseline with active play (obtained with the highest recorded HR), active play resulted in increased latency (p = 0.005), decreased initial velocity (<0.001), decreased time to open (p = 0.019), and decreased lid excursions (p < 0.001).

Table 1. Sample demographics and blink parameters at baseline and active play

Participant Demographics
Number of subjects	44
Age in years ± SD	20.35 ± 0.91
Sex	All male
Previous Concussion, n (%)
0	27 (61.4%)
1	12 (27.3%)
2	3 (6.8%)
3 or more	2 (4.6%)
Blink Parameters	Baseline	Active Play	Mean Difference	Cohen’s D	p-value*
Latency (Msec)	50.39 ± 6.92	53.78 ± 6.86	−3.36 ± 7.66	0.439	0.005
Differential Latency (Msec)	5.02 ± 2.89	5.29 ± 2.59	−0.27 ± 2.91	0.093	0.538
Delta 30 (Msec)	6.60 ± 12.00	5.35 ± 4.14	1.26 ± 12.17	0.104	0.258**
Initial Velocity (pixels/Msec)	4.91 ± 0.98	4.39 ± 1.15	0.52 ± 0.87	0.598	<0.001
Time to Close (Msec)	33.75 ± 4.20	33.48 ± 5.26	0.27 ± 5.45	0.049	0.742
Time to Open (Msec)	283.03 ± 214.28	252.72 ± 192.58	30.30 ± 231.99	0.131	0.019**
Loge Under the Threshold (Msec)	4.64 ± 0.66	4.60 ± 0.63	0.04 ± 0.58	0.069	0.644
Number of Oscillations (qty)	11.42 ± 6.38	12.28 ± 5.54	−0.85 ± 6.40	0.133	0.382
Lid Excursions (pixels)	130.00 ± 24.77	117.85 ± 29.15	12.16 ± 20.65	0.589	<0.001

Note: *p-value from paired t-test; ** p-value from Wilcoxon Signed rank test

Table 2 presents the correlation between the HR and each blink reflex parameter. None of the blink reflex parameters exhibited a direct correlation with the HR, except for the number of oscillations (Person’s r = −0.253, p = 0.013). After separating the active play sessions based on exertion levels, there were 45 scans within the mild exertion-level category and 51 scans in the moderate/higher exertion-level category. Tables 3 shows the groupwise blink parameters at active play relative to the baseline levels. There were no statistically significant groupwise differences between the mild exercise category and the moderate/higher intensity exercise category except oscillations (p = 0.070), which exhibited a trend towards significance (Table 3). Lastly, when comparing changes from baseline between mild and moderate/higher intensity exercise, the number of oscillations (p = 0.017) was the only variable that was significantly different between the two groups (Table 3).

Table 2. Correlation of blink parameters with HR at active play session

Blink Parameter	Correlation Coefficient	p-value*
Latency (Msec)	0.177	0.086
Differential Latency (Msec)	0.177	0.086
Delta 30 (Msec)	−0.080	0.608**
Initial Velocity (pixels/Msec)	−0.001	0.994
Time to Close (Msec)	−0.017	0.873
Time to Open (Msec)	0.007	0.964**
Loge Under the Threshold (Msec)	−0.146	0.158
Number of Oscillations (qty)	−0.253	0.013
Lid Excursions (pixels)	−0.034	0.746

Note: *p-value obtained from Person Correlation; **p-value obtained from Spearman Correlation

Table 3. Blink reflex parameters between mild and moderate/higher intensity levels of HR

	Mild (< 60% HRmax)		Moderate/Higher (≥ 60% HRmax)		p-value*
Number	45		51		-
At active play session (Mean ± SD)
Latency (Msec)	51.32 ± 7.74		52.77 ± 6.87		0.336
Differential Latency (Msec)	4.74 ± 2.38		5.03 ± 2.60		0.563
Delta 30 (Msec)	5.09 ± 3.95		4.66 ± 2.41		0.518**
Initial Velocity (pixels/Msec)	4.29 ± 0.98		4.31 ± 0.91		0.912
Time to Close (Msec)	33.50 ± 5.80		33.39 ± 5.01		0.921
Time to Open (Msec)	287.27 ± 440.89		227.18 ± 154.85		0.364**
Loge Under the Threshold (Msec)	4.66 ± 0.68		4.56 ± 0.54		0.395
Number of Oscillations (qty)	13.15 ± 7.26		10.71 ± 5.75		0.070
Lid Excursions (pixels)	116.83 ± 28.8		114.85 ± 24.70		0.716
Baseline to Active Play Change	Mild		Moderate/Higher
	Mean difference ± SD	Cohen’s D	Mean difference ± SD	Cohen’s D	p-value*
Latency (Msec)	0.78 ± 8.13	0.096	2.48 ± 8.07	0.307	0.324
Differential Latency (Msec)	−0.07 ± 3.24	0.022	0.34 ± 2.78	0.122	0.523
Delta 30 (Msec)	0.75 ± 2.35	0.319	−1.44 ± 11.82	0.122	0.236**
Initial Velocity (pixels/Msec)	−0.63 ± 0.91	0.692	−0.64 ± 1.08	0.593	0.940
Time to Close (Msec)	−0.26 ± 5.26	0.049	−0.36 ± 5.16	0.070	0.928
Time to Open (Msec)	−22.08 ± 479.72	0.046	−50.60 ± 227.10	0.223	0.718**
Loge Under the Threshold (Msec)	−0.018 ± 0.50	0.036	−0.15 ± 0.67	0.224	0.304
Number of Oscillations (qty)	2.34 ± 6.52	0.359	−0.88 ± 6.03	0.146	0.017
Lid Excursions (pixels)	−12.15 ± 19.62	0.619	−14.84 ± 18.17	0.817	0.503

Note: *p-value from independent samples t-test; **p-value from Mann Whitney U test

5. Discussion

The purpose of this study was to assess if blink reflex parameters differed significantly during active play as compared to baseline data. With subjects serving as their own controls, several blink reflex parameters diverged from baseline to active play and included increased latency, decreased initial velocity of eyelids, decreased time to open, and decreased lid excursions. In comparing the two levels of intensity (mild versus moderate/higher), number of oscillations was the only parameter that was significantly different. Prior blink reflex research by Garner et al. (Garner et al., 2018) found that after active play in football, significant differences occurred in latency, differential latency, eyelid excursions, initial velocity of eyelids, time to open, number of oscillations, and total blink time as compared to baseline values. However, the sample size in that study was small (n = 10) and had a lag in post activity data collection. Thus, the purpose of this study was to test a larger cohort and to assess participants at multiple time points during active play, versus at the completion of active play. In comparing the results from both studies, four blink reflex parameters were significantly different with activity and included increased latency, decreased initial velocity of eyelids, decreased time to open, and decreased lid excursions, with all parameters differing 5–10% when compared to their baseline values. For example, latency in the earlier study cited a baseline of 50.2 ms and post value of 55.9 ms, while the current study cited a latency baseline of 50.39 ms and 53.78 ms after activity, both increasing with activity. Number of oscillations and differential latency were the two variables dissimilar between the two studies, with the prior study citing significant changes in these two variables (Garner et al., 2018). A plausible explanation for these differences may be attributed to the timing of the scans. For this study, players provided scans on the sidelines between plays, whereas, in the prior study, players had scans after they had left the field and returned to the locker room, approximately 10–20 minutes after practice. Thus, differences noted in the two studies as it relates to oscillations and differential latency may be due to the timing of scans post-activity, and related to the catecholamine response occurring during and after exercise.

While the catecholamine response during active play was not assessed in this study, evidence suggests that norepinephrine (NE) and dopamine play a role in altering ocular function after prolonged bouts of physical activity (Connell et al., 2017a, 2017b). Connell and colleagues reported significant changes in saccades (defined as oculomotor task) and reflexive eye movements (prosaccades and the quick phase of the optokinetic response) with prolonged exercise (Connell et al., 2017a, 2017b). Increased latency and decreased velocity of the blink reflex in this current study, are comparable with increased saccade latency and decreased saccade velocity (decreased peak velocity and antisaccade peak velocity) cited in the two studies by Connell and colleagues (Connell et al., 2017a, 2017b). Additionally, they found that preventing the inhibition of NE-dopamine reuptake, resulted in protection of these ocular parameters post exercise, thereby establishing the mediating effect of dopamine and NE on ocular function with exercise (Connell et al., 2017a, 2017b). Additionally, an understanding of the blink reflex response with activity requires a comparison of the blink reflex post-head impact, as both conditions put the individual in a state of stress. While it is generally recognized that physical activity is positive stress (eustress) and concussion is a negative stress, it is unknown how these two types of stress affect the blink reflex response. Significant shifts in inflammatory markers and catecholamine responses occur with the acute and chronic phases of concussion (DiBattista et al., 2016; Giza & Hovda, 2014; Hamill et al., 1987; Romeu-Mejia et al., 2019; Rizoli et al., 2017). Rizzoli and colleagues cited a significant increase in NE in an acute response post concussion, with the responses decreasing over 24 h, yet with levels that were significantly higher as compared to controls (Rizoli et al., 2017). Thus, the increased NE response in acute concussion may explain decreased blink reflex latency values cited post-concussion (54.3 ms baseline vs. 51.9 ms post impact, p = 0.028) (Garner et al., 2018), while impaired NE response with prolonged exercise may explain increased latency noted in this study and in the similar ocular function studies (Connell et al., 2017b; Garner et al., 2018).

With the emergence of ocular function assessments in concussion management protocols, it is imperative to evaluate outcomes within the context of the sideline environment, post-physical activity. The outcomes of this study provide data from sideline testing and have meaningful clinical implications for sideline assessment of concussion utilizing blink reflex testing. While previous studies documented blink reflex data from athletes at baseline and post-head impact, there was limited data on the effects of physical activity on blink reflex (Garner et al., 2018). This study aimed to determine the validity of the results obtained in the pilot study (N = 10) (Garner et al., 2018) by evaluating a larger group of subjects and to determine if the differences upheld within a sideline/activity condition. These research findings support the pilot study as it relates to four parameters (latency, initial velocity, time to open, and lid excursion) but found no differences in two parameters (differential latency or number of oscillations), which may be attributed to the differences in scan time between the two studies. The outcomes from this study add to the body of evidence on the unique changes occurring with the blink reflex response as based on one’s body state [rest/homeostasis, healthy active/eustress, and concussed/stress] and provide understanding on the utility of the blink reflex technology in baseline and concussion assessment.

6. Limitations

No baseline HR values were taken prior to the active play data collection. However, each subject was actively monitored at baseline and throughout the pre-season timeframe to evaluate possible health abnormalities that could contribute to an abnormal HR during the testing. No abnormalities were noted during the testing with any of the players, with athletic trainers monitoring the athletes during the scrimmages. Another limitation was the lack of exercise standardization. However, since the goal of this study was to collect data during a regulated scrimmage event and not a lab-simulated session, we were unable to standardize the exercise time or exertion level, which would have been affected by the various positions and player characteristics.

Correction

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Acknowledgments

We would like to acknowledge and thank the football coaches, athletic training staff, and the players at The Citadel for their cooperation and support during the data collection portion of this study.

Disclosure statement

Dr Haider is a paid scientific advisor and statistician for Blinkcns, Inc.

Additional information

Notes on contributors

Dena P. Garner

Dena P. Garner is a professor of exercise science at The Citadel. She received a BA from Furman University and continued her training at The University of South Carolina and Oregon State University in the fields of exercise science and muscle physiology, where she received her MS and Ph.D., respectively. At The Medical University of South Carolina, she obtained a post-doctoral fellowship in the area of neurological effects of experimental autoimmune encephalomyelitis (rat model of multiple sclerosis) within retinal ganglion cells. At The Citadel, her research has focused on the assessment of the blink reflex as a biomarker for neurological health in concussed individuals.