American football is associated with the highest incidence of concussion (11), often used interchangeably with the term mild traumatic brain injury (mTBI), among team sports (13). This is not surprising, given that during play, American football athletes are routinely exposed to head impacts, which result in head rotational acceleration/deceleration forces (inertial loading) and propagation of force through the brain (10). These mechanical forces are known to result in acceleration and deceleration forces on neurons, supporting cells and their projecting fibers, as well as intracellular injury, which leads to diffuse axonal injury (1,8), a characteristic feature of mTBI. Using magnetic resonance imaging in conjunction with diffusion tensor imaging (DTI), Davenport et al. (12) demonstrated that a single season of American high school football resulted in changes in magnetic resonance imaging indicative of axonal injury in the absence of a concussion diagnosis.
Neurofilaments are abundant in key intermediate fibers in neurons and major components of the axonal skeleton (15). Though the exact mechanism of release and subsequent appearance of neurofilament light (NFL) in biological fluids is not known, what is known is that significant changes occur in neurofilaments as a result of diffuse axonal injury as evidenced in animal models of traumatic brain injury (TBI) (for a review, see Sciiedler et al. (33)). Interestingly, elevations in cerebrospinal fluid (CSF) NFL protein have been reported in boxers sustaining concussive or subconcussive head impacts (26,39). The utility of NFL as a marker of axonal damage in humans is further supported by studies in which elevations are reported in the CSF and plasma of those patients suffering from neurodegenerative and neuroinflammation diseases (17,18). Further, a strong correlation has recently been reported between CSF and plasma NFL (18). Like boxing, the sport of American football is physical in nature, resulting in exposure to subconcussive head impacts that vary in magnitude and number over the course of a season, irrespective of level of play (7,10), which have been reported to result in neurophysiological changes (34). Detection of a biomarker, such as NFL, would thus be of significance, given that in the absence of reporting or characteristic display of symptoms, an athlete may play before full recovery, which may further subject the athlete to injury and increase the likelihood of damage to the brain (30).
In their recent position statement, the American Medical Society for Sports Medicine suggested that protective equipment does not reduce the incidence and/or severity of concussion in sport (19) despite advancements (37), highlighting the need for different approaches to prevent brain damage. Nutritional supplementation, specifically docosahexaenoic acid (DHA) (22:6n-3), the principal n-3 long-chain polyunsaturated fatty acid in brain tissue (21) has received considerable attention as a possible intervention to mitigate pathology associated with mTBI. Prior investigation of TBI in animal models has demonstrated that supplemental DHA attenuates axonal damage when administered before insult (6,32). The typical American diet is scarce in both DHA and the n-3 long-chain polyunsaturated fatty acid eicosapentaenoic acid (EPA, 20:5n-3), with an estimated combined intake of about 100 mg·d−1 (14). Although it is known that supplemental DHA results in an increase in plasma DHA in a dose-dependent manner up to approximately 2 g·d−1 which approaches saturation in the average population (5), American football athletes are larger than the average population, and most athletes (38) participate in heavy physical training, which is known to affect fatty acid composition (28). Therefore, larger doses may be necessary.
This study sought to examine changes in serum NFL over the course of the season in American football athletes. We hypothesized that an increase in NFL would occur after those times in which the number and magnitude of impacts increased. Further, given reports of supplemental DHA reducing axonal injury in rodent models of TBI (6,32), we sought to examine the effect of differing doses of supplemental DHA on serum NFL over the course of a season in American football athletes. We hypothesized that American football athletes would require higher doses of supplemental DHA due to their larger size and physical activity and that supplemental DHA would attenuate any increase in serum NFL observed over the course of the season.
This study was conducted according to the Declaration of Helsinki guidelines. All procedures involving human subjects were approved by the institutional review board of Texas Christian University for use of human subjects in research (protocol no. 1404-68-1404). Written consent was obtained from all subjects.
All athlete volunteers were National Collegiate Athletic Association (NCAA) Division I American football athletes cleared to participate in university athletics as determined by the team physician. Exclusion criteria included chronic daily anti-inflammatory (≥20 d) or antihypertensive medication use, medications for blood lipids, fish oil or omega-3 fatty acid supplementation, and consumption of more than two servings of fish per week. Athletes injured or unable to participate in regularly schedule conditioning or competitions were also excluded. A consort diagram is provided in Figure 1, outlining reasons for dropout and/or exclusion. A total of 81 athletes completed the study.
A randomized, double-blind, placebo-controlled, parallel design study was used to examine the effect of differing doses of DHA on plasma fatty acids and serum NFL over the course of a season (189 d) of American football to include off-season summer conditioning (57 d), preseason camp (23 d), and the competitive season (109 d). Before the start of summer conditioning, the American football athletes were randomly assigned to ingest an oil mixture providing either 2 g·d−1 DHA, 4 g·d−1 DHA, 6 g·d−1 DHA, or placebo, that is, corn oil, for the study duration. Before the start of the season, blood was sampled at specific intervals coincident with changes in intensity, hours of contact, and coincident with those times in which the number and magnitude of head impacts have been reported to change (9,10). Upon reporting for summer conditioning, a baseline blood sample was taken when volunteers returned from a period of no contact (9 wk) (baseline; T1). Summer conditioning consisted of strength and conditioning workouts Monday through Friday for approximately 2 h·d−1. A second sample was obtained at the conclusion of summer training before preseason training camp (precamp; T2). In the time between baseline and the start of training camp, no head-to-head contact took place, but many of the athletes elected to participate in nonsupervised seven on seven practices, which involve no contact or equipment and occurred approximately twice per week for 1 h. A third sample was collected after preseason training camp (postcamp; T3). Camp began with three noncontact practices in shorts and helmets, one practice in helmet and shoulder pads, followed by full gear practices. The first 5 d, practices were held once per day. Thereafter, two-a-day practices were held every other day for the remainder of camp. One day per week (Sunday), the athletes were off from training and practice. The remaining blood samples were taken throughout the competitive season on the Monday after a Saturday game (48 h) (T4–T8). The first two samplings (T4, T5) occurred during preconference play, whereas the last three (T6–T8) occurred during conference play. During the competitive season, athletes underwent a 1-h shorts and helmets practice on Sunday, Monday was an off day, followed by full pad practice on Tuesday, helmets and shoulder pads on Wednesday, Thursday, and Friday were shorts and helmets only (no full contact, incidental only). During the competition season, tackles were not taken to the ground, hit and wrap only. Additionally, athletes would perform three strength and conditioning sessions per week. The maximum (and median) number of head impacts per season for an NCAA American football athlete has been reported to range from 15.6 to 24 (4.8–7.5) and 58.5 to 86.1 (12.1–16.3) for practices and games, respectively (9). All samplings throughout the season occurred within 14- to 28-d intervals, with no sampling occurring more than 28 d from the previous one. Given that some American football athletes are known to have a higher number of total contact hours due to the number of plays (repetitions) per game, serum NFL was compared across starters and nonstarters. Starters were defined as athletes known to go out with the first or second team, first or second on the depth roster, and take a majority of the repetitions (∼20 to 40+ per game).
Before the beginning of summer training, a 130-man roster of prospective volunteers from a NCAA Division I American football team was provided to research personnel. Potential participants were randomized by player position (quarterback, offensive line, running back, tight end, wide receiver, defensive line, linebacker, cornerback, safety, specialist [i.e., kicker], incoming freshmen—no positions listed) provided by coaching staff and subsequently randomized to one of four treatment groups (2 g·d−1 DHA, 4 g·d−1 DHA, 6 g·d−1 DHA, or placebo) using a random number generator. The algal DHA oil used in this study (DHA-S oil; DSM Nutritional Products, Columbia, MD) was derived from Schizochytrium sp. containing 35%–45% DHA by weight. Wesson® corn oil, containing < 0.1% DHA, was used as the placebo and in the 2 g·d−1 and 4 g·d−1 DHA oil mixtures to ensure equal volume. To ensure proper dose the lower limit (35%) of DHA content by weight contained in the DHA-S oil was used for calculations. The oil mixture was weighed to the nearest 0.1 g on food scales (Ohaus CS-200, Parsippany, NJ). All mixtures were flavored with artificial flavoring as a masking agent. Stability of oil mixture was verified by procedures outlined in “Fatty acid composition of plasma” for preparation up to 7 d before delivery. The fatty acid composition for both oils is shown in Table 1.
Participants were advised as to which foods were high in DHA and asked to limit servings to no more than two per week during the study. Supplement adherence was monitored daily via visual supervision by the same research personnel. To ensure compliance, supplement was given only on days when the participants reported to the training facility resulting in participants receiving supplement for 5, 7, and 5 d·wk−1 for summer training, preseason camp, and the competitive season, respectively. Supplement compliance was set at ≥80% for each individual over the course of the study. The days when supplement was not given were not taken into account for calculating compliance.
Blood sampling and preparation.
The night before each blood collection, participants were verbally reminded to ingest only water after 2200 h. On the day of blood sampling, participants reported to the athletic training facility after an overnight fast (≥ 8 h), and supine blood samples were collected via venipuncture from the antecubital fossa region using standard, sterile phlebotomy procedures. Blood samples were collected in spray-coated K2 ethylenediaminetetraacetic acid vacutainer tubes and serum vacutainer tube with no additive (BD Diagnostics, Franklin Lakes, NJ). Vacutainer tubes were kept on ice before blood collection and immediately placed on ice after blood collection. All samples were centrifuged at 2000g for 30 min at 4°C (Beckman Coulter, Allegra X-15R, Brea, CA) within 30 min of collection. Aliquots of serum, plasma, and red blood cells, collected via sterile pipette from center of the red blood cell pack within the ethylenediaminetetraacetic acid vacutainer tubes, were immediately transferred to prelabeled polypropylene vials .
Serum NFL levels were determined using the NF-Light kit from UmanDiagnostics (UmanDiagnostics, Umeå, Sweden), transferred onto the Simoa platform using a homebrew kit (Quanterix Corp, Boston, MA). The lower limit of quantification was 1.95 pg·mL−1. The analyses were performed by a board-certified laboratory technician in one round of experiments using one batch of reagents with intra-assay coefficients of variation below 10%.
Fatty acid composition of plasma.
Total lipid was extracted and methylated according to previously described procedures (2). Fatty acid methyl esters were analyzed with a Varian gas chromatograph (model CP-3800 fixed with a CP-8200 autosampler; Varian Inc., Walnut Creek, CA) equipped with a fused silica capillary column CP-Sil88 (100 m × 0.25 mm i.d.; Chrompack Inc., Middleburg, The Netherlands) (16). Individual fatty acid methyl esters were identified using genuine standards (Nu-Chek Prep, Inc., Elysian, MN), and response factors were calculated from a library of commercially available methyl esters of palmitic, oleic, linoleic, and arachidonic acid (ARA) from the same manufacturer. Data are expressed as percent (%) total fatty acids.
A priori power analysis was conducted using G*Power version 3.1.9 to determine the minimum sample size required to find significant changes in the proportion of DHA in total plasma fatty acids with a desired level of power set at 0.80, an α-level at 0.05, and a standardized effect size calculated from a previous pilot study. It was determined that a total of 20 subjects (five per group) were needed to ensure adequate power.
The effects of supplementation and time on variables of interest were calculated from a mixed model ANOVA in SPSS V.22 (IBM Corporation; Armonk, NY). From that model, estimates and uncertainty of the large-sample effect size for the effect of treatment and time on dependent measures were derived from the model to allow a magnitude-based approach to inference (20). Effect sizes were calculated and a modified classification system (trivial, 0.0–0.2; small, 0.2–0.6; moderate, 0.6–1.2; large, 1.2–2.0; very large, >2.0; extremely large, >4.0) was used to further interpret the magnitude of change. The probability (likelihood) that a contrast was at least greater than the smallest threshold, calculated as the standardized change of 0.2 times the between-subject SD at baseline among all treatments was qualified as follows: 0.5%, almost certainly not; 0.5%–5%, very unlikely; 5%–25%, unlikely; 25%–75%, possible; 75%–95%, likely; 95%–99.5%, very likely; and 99.5%, almost certain. In the case where the majority (>50%) of the confidence interval lies between the threshold for substantiveness and the probability of benefit or harm is <5.0%, the effect was qualified trivial with the appropriate likelihood qualifier (20).
Compliance and reported side effects.
Mean compliance over the course of the 189-d intervention was 94%, 90%, 95%, and 91% for 2 g·d−1 DHA, 4 g·d−1 DHA, 6 g·d−1 DHA, and placebo; respectively. Side effects of supplemental DHA included initial gastrointestinal distress (n = 4), poor palatability (n = 1), and smell (n = 1). These were only reported once over the duration of the study.
Effect of supplemental DHA on plasma DHA.
The change from T1 to T8 in proportion DHA, EPA, and ARA of total plasma fatty acids is shown in Figures 2A–C, respectively. DHA supplementation increased the proportion of DHA in plasma fatty acids from baseline by a most likely (99.9%) large magnitude (ES = 1.74–2.17) in a dose-dependent manner. Clear contrasts were observed when examining change from baseline in each dose, with very likely (98.0%) and most likely (99.6%) substantially greater delta in the 6 g·d−1 DHA treatment group, when compared with 4 g·d−1 (ES = 0.7; P = 0.006) and 2 g·d−1 (ES = 1.0; P = 0.001). Supplemental DHA also increased plasma EPA in the 4 g·d−1 (ES = 0.6; P = 0.130) and 6 g·d−1 (ES = 1.7; P < 0.001) treatment groups. However, clear contrasts for EPA were only observed in the 6 g·d−1 dose (mean difference, 0.23; 90% confidence interval, ±0.15%) when change from baseline was compared with placebo (ES = 0.9; P = 0.016). ARA decreased in all DHA treatment groups, as expected, with the most substantial change observed in the 2 g·d−1 and 4 g·d−1 treatment groups (ES = 1.2; P = 0.001 and ES = 1.2; P = 0.002, respectively).
Serum NFL over the course of the season.
Serum NFL increased substantially over the course of the season in those participants categorized as starters (Figure 3A). The observed increase occurred coincident with changes in intensity and hours of contact. The postcamp (T3) serum NFL increase observed was very likely (97.4%), but small in magnitude in starters (ES = 0.4; P = 0.001) compared with baseline (T1). As expected, a similar albeit smaller increase was observed in nonstarter (ES = 0.3; P = 0.043) as those in this category would have experienced the most contact during the time from precamp (T2) to postcamp (T3). All other contrasts in the nonstarters were trivial. However, serum NFL continued to be elevated the remainder of the competition season in starters, with a further very likely (99.5%) increase of moderate magnitude occurring at T6 (ES = 0.6; P < 0.001), coincident with conference play. A further increase was noted at T7 (ES = 0.8; P < 0.001), which remained elevated to the end of the competitive season (ES = 0.7; P < 0.001). The increase observed over the course of the study in starters resulted in substantial differences noted over the course of the study when compared with nonstarters (Table 2). Similarly, the area under the curve calculation determination resulted in a very likely (97.4%) difference of moderate magnitude (600; ± 430 pg·mL−1; ES = 0.6; P = 0.024) between starters (1995 ± 1383 pg·mL−1) and nonstarters (1398 ± 581 pg·mL−1).
Effect of supplemental DHA attenuates on serum NFL.
Because of the lack of change in serum NFL, those athletes categorized as nonstarters were excluded from further analyses. The effect of DHA supplementation irrespective of dose (i.e., collapsed across treatments) compared with placebo is presented in Figure 3B, whereas Figure 3C shows the effect of the different doses on serum NFL over the course of the football season. Both figures are presented as percent change from baseline. Statistical contrasts comparing each dose to placebo with corresponding mean ± SD are shown in Table 3. When collapsed across all treatment groups, supplemental DHA likely (92.3%) attenuated serum NFL postcamp (P = 0.070; Table 3), coincident with postcamp (T3), in which a substantial increase in serum NFL was observed in starters. During conference play, when a second substantial increase in serum NFL was observed among starters, supplementation with DHA resulted in a likely (87.1%) lowering effect compared with placebo (P = 0.144, T6), which continued through the end of the study when a likely (T7, 97.5%; T8, 98.9%) lowering of serum NFL was observed in those supplementing with DHA compared with placebo (P = 0.022, T7; P = 0.012, T8). Examination of the effect of the differing doses on serum NFL resulted in fewer athletes in each group (Table 3). However, we observed a likely and very likely substantial lowering effect of 2 g·d−1 DHA compared with placebo over the course of the study. In those supplementing with 4 g·d−1 DHA and 6 g·d−1 DHA, all contrasts were unclear with the exception of the final time point in which a likely (91.0%) substantial lowering effect was observed in those in the 4 g·d−1 treatment.
In this study, we examined the effect of differing doses of DHA on plasma fatty acids and serum NFL, a biomarker of head trauma, over the course of a season in American football athletes. This study provides novel data specific to the optimal dose of DHA for plasma and biomarker changes in American football athletes. Further, the data presented herein demonstrate that a season of American football is associated with some level of subconcussive injury, which results in a measurable increase in a marker of axonal damage. Most importantly, we report for the first time that supplemental DHA, irrespective of dose, may likely attenuate elevations in serum NFL coincident with those times in which an increased number and magnitude of head impacts typically occurs in American football athletes.
The neuroprotective effects of supplemental DHA observed in rodent models of TBI demonstrated greatest efficacy when administered at a dose of 40 mg·kg−1·d−1) (6), which corresponds to a dose of approximately 4 g·kg−1·d−1 in today’s collegiate American football athlete (29). However, in analyzing the dose–response effect, Arterburn et al. (5) reported a dose-dependent relationship in which plasma DHA increases up to a dosage of approximately 2 g·d−1; above this dose, the authors reported only incremental increases are observed as saturation is approached (5). The finding of a dose–response effect is in agreement with those findings (5); however, the substantial increases observed at the 4 g·d−1 and 6 g·d−1 dose suggest that higher doses may be necessary because no apparent plateau was noted. This is not surprising given the large size of American football athletes compared with the general population and other athletes (38). Despite having relatively high body mass, the subjects in those studies analyzed by Arterburn et al. (2) were lower than those reported in American football athletes (23,29). Further, American football athletes undergo rigorous physical training and activity, which is known to influence plasma fatty acid composition (28).
The lack of an observable increase in EPA in those receiving 2 g·d−1 DHA combined with a lack of clear contrasts when comparing change from baseline in those receiving 4 g·d−1 DHA versus placebo supports the assertion that a higher dose may be necessary in American football athletes. While retroconversion of DHA to EPA is regularly observed (5,25), Mori et al. (24) reported only a small nonsignificant increase in EPA following a 6-wk intervention in which overweight men received supplemental DHA in a dose of 4 g·d−1 DHA. Further, physical activity is known to affect fatty acid composition (28) which may be why increases in EPA are not readily observed in this population. Interestingly, the largest decrease in ARA occurred in those receiving 2 g·d−1 DHA. Though a similar dose-dependent reduction in ARA is typically observed, the response has been reported to be variable when examining a number of studies (5). This study and in particular changes in the fatty acid profile contributes important new information specific to optimizing DHA supplementation in American football athletes.
It is well known that American football athletes are exposed to head impacts that vary in number and magnitude over the course of a season (9). Further, those routine head impacts sustained result in rotational acceleration/deceleration and propagation of force through the brain (10), impacts known to cause diffuse axonal injury (1,8). It has been suggested that NFL, albeit in CSF, is one of the most sensitive biomarkers of head trauma, specifically axonal injury (26,40). Recently, Al Nimer et al. (3) reported that a significant relationship exists between CSF NFL and serum NFL in severe TBI. As such, elevations of NFL in peripheral blood would indicate some level of axonal injury. Thus, the finding of marked elevations after periods in which an increase in number and magnitude of head impacts was likely (9) suggests that some level of axonal damage occurs in American football athletes over the course of a season in the absence of a concussion diagnosis.
Indeed, Davenport et al. (12) reported that the risk weighted exposure, a metric defined as the collected risk of concussion over the course of a season(36), showed strong associations with DTI scalars in American football athletes at the high school level in the absence of a concussion diagnosis. Most studies using DTI in TBI and mTBI report that the changes reported in those specific scalars identified by Davenport et al. (12) result from axonal injury (4,27). The number of high-level impacts are more frequent in collegiate compared with high school players (31); thus, our data support that the subconcussive impacts sustained by collegiate American football players results in some degree of axonal injury. The primary inferential limitation to the current data is the lack of an outcome measure related to the actual number of head impacts sustained over the course of the season. As a result, our ability to discern if the elevations in serum NFL were a result of axonal damage caused by head impacts or from another source, such as muscle, is limited. However, despite participating in practice over the course of the season, serum NFL did not vary in those categorized as nonstarters, suggesting that the elevations observed herein were in fact a result of impacts to the head.
Perhaps, the most novel finding of the current study was that supplemental DHA likely attenuated the increase observed in serum NFL. Interestingly, the lowest dose (2 g·d−1) appeared to produce the most marked reductions in serum NFL compared with placebo. However, inference about the differing doses is limited by the number of starters in each group. Previous studies in rodent models of TBI support the possibility that DHA reduces markers of axonal injury (6,32). Fewer numbers of β-amyloid precursor protein–positive axons, a marker of axonal injury, were reported in animals receiving supplemental DHA for 30 d before impact acceleration injury (6). More recently, Schober et al. (32) examined the effect of DHA on DTI indices of white matter injury using an established model for pediatric TBI, controlled cortical impact (22). DHA decreased the increment in radial diffusivity observed in the DTI following controlled cortical impact, increased radial diffusivity is correlated with histologic axonal damage (35). Thus, our data are in support of previous studies in which DHA has been shown to provide neuroprotective effect in rodent models of TBI. Although a number of mechanisms have been proposed, the exact mechanism underlying the neuroprotective effects of DHA are as of yet unknown. Our findings are in support of those findings in which DHA attenuated markers of axonal injury in rodent models. However, we report on the potential effectiveness of DHA for axonal injury caused by subconcussive impacts in American football athletes, the first to date.
The current study has several limitations. First, our results are limited by the sample size, particularly because it relates to starters. Based on preliminary data and the absence of an optimal dose regimen for DHA for the larger American football athlete at the time of design, we sought to experimentally define an optimal dose. In doing so, this limited the number of starters in each treatment group, making interpretation of the optimal dose for neuroprotective effect of DHA difficult. However, the likely decrease in serum NFL in the presence of all doses of DHA (collapsed across treatment) was apparent throughout the season, though this may have been attributed mostly to those athletes receiving 2 g·d−1 DHA. A larger sample size in a confirmatory study is suggested. Second, the lack of an outcome measure for number of head impacts limits inference as to the origination of NFL appearing in peripheral blood. No increase was observed in the nonstarters participating in strenuous practice which suggests that NFL was indeed from head impacts. Further study is warranted to examine the relationship between the number of head impacts and serum NFL levels in this population using a more direct measure of head impacts (i.e., telemetry data collection). The inclusion of DTI would also provide additional support.
In conclusion, the data presented herein demonstrate that American football athletes likely require a higher dose of DHA than the average population. Further, serum NFL increases in those categorized as starters coincident with those times in which the number of head impacts likely increases. Most importantly, we report for the first time that supplemental DHA likely attenuates the increase in serum NFL, suggesting a neuroprotective effect of DHA, specifically because it relates to axonal injury, the central pathogenic mechanism in mTBI. The latter agrees with previous reports of the neuroprotective effects of DHA in rodent models of TBI. Our study contributes important information on the effects of DHA on a biological marker of head trauma in a population known to sustain a significant number of head impacts. Given the potential of DHA to provide neuroprotection and the fact that it is well tolerated and safe at any age, further study is warranted to elucidate the true nature of the effect and its potential prophylactic use.
Supplements were provided by DSM Nutritional Products. Support for this study was received from the following: DSM Nutritional Products, Office of the Provost for Texas Christian University, the College of Education and Human Development at George Mason University, Swedish Research Council, VINNOVA, the Knut and Alice Wallenberg Foundation, Frimurarestiftelsen.
K. B. and H. Z. are cofounders of Brain Biomarker Solutions in Gothenburg AB, a GU Venture-based platform company at the University of Gothenburg.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
The authors would like to thank the Texas Christian University athletes, coaches, and sports medicine staff as well as the Texas Christian University Kinesiology Department graduate students and faculty for their help. Further, the authors would like to thank Dr. Stephen Smith and Terri Blackmon from Texas A&M University and Dr. Michael Lewis of the Brain Health Education and Research Institute.
1. Adams JH, Doyle D, Ford I, Gennarelli T, Graham D, McLellan D. Diffuse axonal injury in head injury: definition, diagnosis and grading. Histopathology
. 1989; 15(1): 49–59.
2. Adams TH, Walzem RL, Smith DR, Tseng S, Smith SB. Hamburger high in total saturated and trans-fatty acids decreases HDL cholesterol and LDL particle diameter, increases plasma TAG, in mildly hypercholesterolaemic men. Br J Nutr
. 2010; 103: 91–8.
3. Al Nimer F, Thelin E, Nyström H, et al. Comparative assessment of the prognostic value of biomarkers in traumatic brain injury reveals an independent role for serum levels of neurofilament light. PLoS One
. 2015; 10(7): e0132177.
4. Arfanakis K, Haughton VM, Carew JD, Rogers BP, Dempsey RJ, Meyerand ME. Diffusion tensor MR imaging in diffuse axonal injury. Am J Neuroradiol
. 2002; 23(5): 794–802.
5. Arterburn LM, Bailey Hall E, Oken H. Distribution, interconversion, and dose response of n-3 fatty acids in humans. Am J Clin Nutr
. 2006; 83(suppl): 1467S–76.
6. Bailes JE, Mills JD. Docosahexaenoic acid reduces traumatic axonal injury in a rodent head injury model. J Neurotrauma
. 2010; 27: 1617–24.
7. Broglio SP, Schnebel B, Sosnoff JJ, et al. The biomechanical properties of concussions in high school football. Med Sci Sports Exerc
. 2010; 42(11): 2064.
8. Browne KD, Chen X-H, Meaney DF, Smith DH. Mild traumatic brain injury and diffuse axonal injury in swine. J Neurotrauma
. 2011; 28(9): 1747–55.
9. Crisco JJ, Fiore R, Beckwith JG, et al. Frequency and location of head impact exposures in individual collegiate football players. J Athl Train
. 2010; 45(6): 549.
10. Crisco JJ, Wilcox BJ, Beckwith JG, et al. Head impact exposure in collegiate football players. J Biomech
. 2011; 44(15): 2673–8.
11. Daneshvar DH, Nowinski CJ, McKee A, Cantu RC. The epidemiology of sport-related concussion. Clin Sports Med
. 2011; 30(1): 1–17.
12. Davenport EM, Whitlow CT, Urban JE, et al. Abnormal white matter integrity related to head impact exposure in a season of high school varsity football. J Neurotrauma
. 2014; 31(19): 1617–24.
13. DeKosky ST, Blennow K, Ikonomovic MD, Gandy S. Acute and chronic traumatic encephalopathies: pathogenesis and biomarkers. Nat Rev Neurol
. 2013; 9(4): 192–200.
14. Ervin RWJ, Wang C, Kennedy-Stephenson J. Dietary intake of fats and fatty acids for the United States population: 1999–2000. Adv Data Vital Health Stat
. 2004; 348: 1–6.
15. Fletcher DA, Mullins RD. Cell mechanics and the cytoskeleton. Nature
. 2010; 463(7280): 485–92.
16. Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem
. 1957; 226(1): 497–509.
17. Gaiottino J, Norgren N, Dobson R, et al. Increased neurofilament light chain blood levels in neurodegenerative neurological diseases. PLoS One
. 2013; 8(9): e75091.
18. Gisslén M, Price RW, Andreasson U, et al. Plasma concentration of the neurofilament light (NFL) protein is a biomarker of CNS injury in HIV infection: a cross-sectional study. EBioMedicine
. 2016; 3: 135–40.
19. Harmon KG, Drezner JA, Gammons M, et al. American Medical Society for Sports Medicine position statement: concussion in sport. Br J Sports Med
. 2013; 47(1): 15–26.
20. Hopkins W, Marshall S, Batterham A, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc
. 2009; 41(1): 3.
21. Kim HY. Novel metabolism of docosahexaenoic acid in neural cell. J Biol Chem
. 2007; 26: 18661–5.
22. Lighthall JW. Controlled cortical impact: a new experimental brain injury model. J Neurotrauma
. 1988; 5(1): 1–15.
23. Melvin MN, Smith-Ryan AE, Wingfield HL, Ryan ED, Trexler ET, Roelofs EJ. Muscle characteristics and body composition of NCAA division I football players. J Strength Cond Res
. 2014; 28(12): 3320–9.
24. Mori TA, Bao DQ, Burke V, Puddey IB, Beilin LJ. Docosahexaenoic acid but not eicosapentaenoic acid lowers ambulatory blood pressure and heart rate in humans. Hypertension
. 1999; 34(2): 253–60.
25. Mori TA, Burke V, Puddey IB, et al. Purified eicosapentaenoic and docosahexaenoic acids have differential effects on serum lipids and lipoproteins, LDL particle size, glucose, and insulin in mildly hyperlipidemic men. Am J Clin Nutr
. 2000; 71(5): 1085–94.
26. Neselius S, Brisby H, Theodorsson A, Blennow K, Zetterberg H, Marcusson J. CSF-biomarkers in Olympic boxing: diagnosis and effects of repetitive head trauma. PLoS One
. 2012; 7(4): e33606.
27. Newcombe V, Williams G, Nortje J, et al. Analysis of acute traumatic axonal injury using diffusion tensor imaging. Br J Neurosurg
. 2007; 21(4): 340–8.
28. Nikolaidis MG, Mougios V. Effects of exercise on the fatty-acid composition of blood and tissue lipids. Sports Med
. 2004; 34(15): 1051–76.
29. Oliver JM, Lambert BS, Martin SE, Green JS, Crouse SF. Predicting football players’ dual-energy x-ray absorptiometry body composition using standard anthropometric measures. J Athl Train
. 2012; 47(3): 257.
30. Saunders RL, Harbaugh RE. The second impact in catastrophic contact-sports head trauma. JAMA
. 1984; 252(4): 538–9.
31. Schnebel B, Gwin JT, Anderson S, Gatlin R. In vivo study of head impacts in football: A comparison of National Collegiate Athletic Association Division I versus high school impacts. Neurosurgery
. 2007; 60(3): 490–6.
32. Schober ME, Requena DF, Abdullah OM, et al. Dietary docosahexaenoic acid Improves cognitive function, tissue sparing, and magnetic resonance imaging indices of edema and white matter injury in the immature rat after traumatic brain injury. J Neurotrauma
. 2015; 32: 1–13.
33. Sciiedler DG, Chuah MI, Kirkcaldie MT, Vickers JC, King AE. Diffuse axonal injury in brain trauma: insights from alterations in neurofilaments. Front Cell Neurosci
. 2014; 8.
34. Talavage TM, Nauman EA, Breedlove EL, et al. Functionally-detected cognitive impairment in high school football players without clinically-diagnosed concussion. J Neurotrauma
. 2014; 31(4): 327–38.
35. Tuor UI, Morgunov M, Sule M, et al. Cellular correlates of longitudinal diffusion tensor imaging of axonal degeneration following hypoxic–ischemic cerebral infarction in neonatal rats. Neuroimage: Clin
. 2014; 6: 32–42.
36. Urban JE, Davenport EM, Golman AJ, et al. Head impact exposure in youth football: high school ages 14 to 18 years and cumulative impact analysis. Ann Biomed Eng
. 2013; 41(12): 2474–87.
37. Viano DC, Halstead D. Change in size and impact performance of football helmets from the 1970s to 2010. Ann Biomed Eng
. 2012; 40(1): 175–84.
38. Yamamoto JB, Yamamoto BE, Yamamoto PP, Yamamoto LG. Epidemiology of college athlete sizes, 1950s to current. Res Sports Med
. 2008; 16: 111–27.
39. Zetterberg H, Hietala MA, Jonsson M, et al. Neurochemical aftermath of amateur boxing. Arch Neurol
. 2006; 63(9): 1277–80.
40. Zetterberg H, Smith DH, Blennow K. Biomarkers of mild traumatic brain injury in cerebrospinal fluid and blood. Nat Rev Neurol
. 2013; 9(4): 201–10.