Research Article, J Athl Enhancement Vol: 4 Issue: 2
Neck Muscle Strength Training in the Risk Management of Concussion in Contact Sports: Critical Appraisal of Application to Practice
| Ian Gilchrist1,2, Michael Storr3, Elizabeth Chapman4 and Lucie Pelland1,2* | |
| 1School of Rehabilitation Therapy, Queen’s University, Kingston, Canada | |
| 1The Human Mobility Research Centre at Queen’s University and Kingston General Hospital, Kingston, Canada | |
| 3Kingston General Hospital, Department of Pediatrics, Kingston, Canada | |
| 4BTE Technologies Inc., Milton, Ontario, Canada | |
| Corresponding author : Lucie Pelland, PT, PhD Associate Professor, Queen’s University, School of Rehabilitation Therapy, Louise D. Acton Building, Kingston, Ontario, Canada, K7L 3N6 Tel: +1-613-533-3237 E-mail: Lucie.Pelland@queensu.ca |
|
| Received: July 03, 2014 Accepted: June 02, 2015 Published: June 09, 2015 | |
| Citation: Gilchrist I, Storr M, Chapman E, Pelland L (2015) Neck Muscle Strength Training in the Risk Management of Concussion in Contact Sports: Critical Appraisal of Application to Practice. J Athl Enhancement 4:2. doi:10.4172/2324-9080.1000195 |
Abstract
Neck Muscle Strength Training in the Risk Management of Concussion in Contact Sports: Critical Appraisal of Application to Practice
Background: Neck strength training has been advocated as a player-specific modifiable factor in the risk management for concussion in contact sports. A scoping review of the literature was undertaken to address two specific aims. The first was to identify and critically appraise the level and quality of evidence relating neck strength and resistance training to concussion incidence and risk in contact sports. The second was to compare and contrast the effectiveness of resistance neck strengthening programs and to evaluate effects of increased strength in attenuating the postimpact kinematics of the head, a proxy measure of concussion risk. Methods: Structured search of five electronic databases (Ovid MEDLINE, CINAHL, PubMED, EMBASE, and AMED), combining MeSH and generic search terms relating neck strength to concussion biomechanics, risk and incidence. Level of research evidence (Oxford Centre of Evidence-based Medicine) and methodological quality were determined (PEDro and Newcastle-Ottawa Scales). Results: Total isometric neck strength predicted concussion incidence in one prospective study (level 1b). The effect size of strength on concussion incidence was small (Cohen’s d, 0.29). Peak isometric strength did not predict the odds of sustaining a moderate or severe head impact in contact sports (level 1b, 2b, and 4). Short-latency anticipatory strength exerts an attenuating effect on post-impact kinematics of the head (level 1b, 2b) and can be facilitated by selective parameters of isotonic strength training. Methodological quality of the research evidence ranged from 6/10 to 8/10 for controlled trials and 6/9 to 9/9 for case-series and cohort studies. Conclusion: Short-latency strength, developed prior to impact, is a key modifying variable of the post-impact kinematics of the head. By facilitating short-latency neck strength, muscle strength training is a potential target to favorably influence concussion risk, but further study is required to determine the translation of neck/head kinematics to concussion risk. Standardized methods for assessment of multi-directional short-latency, and peak neck, strength need to be adopted and combined with prospective studies.
Keywords: Concussions; Neck strength; Resistance training; Post-impact head kinematics; Concussion risk; Neck stiffness
Keywords |
|
| Concussions; Neck strength; Resistance training; Post-impact head kinematics; Concussion risk; Neck stiffness | |
Abbreviations |
|
| HN: Head-Neck; PEDro: Physiotherapy Evidence Database; RCT: Randomized Controlled Trial; Non-RCT: Non-randomized Controlled Trial; NOS: Newcastle-Ottawa Scale; MDC95%: Minimum detectable change with 95% confidence; lbs: pound; Kg: kilogram; N: Newton; MADYMO: Mathematical Dynamic Model; HIC: Head Injury Criterion; ACSM: American College of Sports Medicine; CI: Confidence Interval; SCM: sternocleidomastoid; UFT: Upper Fibers of Trapezius; RFD: Rate of Force Development; EMG: Electromyography; RM: Repetition Maximum; NCAA: National Collegiate Athletic Association | |
Introduction |
|
| In response to increasing evidence of the severity of acute effects of concussion on neurocognitive function and of the possibility for their lasting impairments on health [1-3], implementation of risk management strategies for concussion has become a priority for sports governing bodies [4-7]. Effective risk management requires a multi-factorial approach, with athlete preparation and sport readiness being fundamental components [8]. Within the context of contact sport, strength training of the neck musculature has increasingly been advocated as a player-specific modifiable factor to lower the odds of sustaining a concussion [9-14]. As stronger muscles generate higher peak magnitudes of isometric tension at faster rates of force development [15], it is postulated that strength training of the neck musculature would enhance the early resistance of the head and neck (HN) segment to externally applied forces, attenuating the post-impact kinematic response of the head and, thereby, lowering the risk for concussion [9,10]. | |
| While this basic research on muscle mechanics provides theoretical support for neck strengthening programs that are being promoted as preventative measures for concussions in contact sports [16-18], the research evidence specifically relating neck strength to concussion risk, incidence and severity has yet to be comprehensively evaluated. Therefore, a scoping review of the literature was undertaken to address two specific aims. The first was to identify and critically appraise the level and quality of evidence relating neck strength and resistance training of the neck musculature to the incidence of concussion in contact sports. The second was to compare and contrast the effectiveness of resistance training programs in producing absolute gains in isometric neck strength in non-clinical populations, and to evaluate effects of increased strength in attenuating the postimpact kinematics of the head, which provides a proxy measure of concussion risk. | |
Methods |
|
| The scoping review was performed using the methods outlined by Arksey et al. [19] and Anderson et al. [20]. Five databases were searched - Ovid MEDLINE, CINAHL, PubMED, EMBASE, and AMED – using two structured search strategies. The first strategy combined MeSH and generic terms relating neck strength, measured at baseline or following a resistance training intervention, to concussion risk and incidence, and to concussion-relevant kinematics of the HN segment. The second search focused on outcomes of neck training programs and the relationship of these outcomes to the kinematics of the HN segment. The search strategies are described in Table 1 and 2, and the searches are up to date to January 2015, week 4. In agreement with the scoping nature of the review, the search was not limited by study design; all experimental and quasi-experimental designs, and systematic reviews outlined by the Oxford Center for Evidence-based Medicine were included. | |
 |
Table 1: MeSH Headings and Keywords for Search on Neck Strength and Concussion Biomechanics and Risk. |
|
Table 2: Medical Subject Headings (MeSH) and Keywords for Search on Resistance Training for the Neck Musculature in Non-Clinical Populations. |
| Search outcomes | |
| The first search identified 343 articles relating neck strength and concussion incidence and risk (Table 1). Of these, 46 titles were redundant, leaving 297 studies for abstract review. Another 262 studies were excluded at this phase of the review process as neck strength was either evaluated within the context of intervention studies in clinical populations or concussion risk and incidence were not explicitly measured outcomes. Of the thirty-five remaining studies, twelve were general review articles that did not include either original or systematically reviewed data, and three studies could not be retrieved. Five additional studies were identified by manual search of the reference list of retrieved studies and by Google Scholar alerts of new articles on concussion. Of this final set of twenty-five articles, twelve were excluded following full review as they provided a general context for interpreting research evidence on concussion but did not contribute specific data relating neck strength to concussion incidence, risk, or post-impact kinematics of the HN segment. Therefore, thirteen unique articles were included in the critical appraisal of evidence. | |
| The second search identified 174 articles on resistance training programs for the neck musculature (Table 2). Of this original set, four redundant titles were excluded and one article could not be retrieved. Abstracts were reviewed for the remaining 169 studies, with 161 being excluded at this phase as they evaluated the effectiveness of strength training programs of the neck and shoulder girdle in relation to the incidence of neck pain in healthy populations, comparatively evaluated the outcomes of different training modalities on strength using repeated measures analysis within a single session, or focused on outcomes between healthy controls and clinical populations. One general review article was also excluded, as it did not present either original or systematically reviewed data. Three additional strength training studies were identified through manual search of the reference list of retrieved studies, resulting in ten resistance training programs included in the critical appraisal of effectiveness. | |
| Data analysis | |
| The guidelines of Law and MacDermid [21] were used to appraise retrieved studies; summaries of experimental design and methods, statistical comparison, measured outcome and findings were provided for all studies included in the analysis (Tables 3-6). The Physiotherapy Evidence Database (PEDro) scale was used to evaluate the methodological quality of experimental controlled trials with and without randomization (RCT and non-RCT) [22,23], while the Newcastle-Ottawa Scale (NOS) was used to evaluate the quality of case-series and cohort studies [24]. On the PEDro scale, the criterion for high quality methodology is a score ≥ 6/10, with a maximum score of 8/10 possible for non-RCTs. The NOS provides a continuous grading of methodological quality for cohort and case-series studies from 0 to 9, with no definition of cut-off criteria to define high quality methodology. The level of research evidence was determined using the Oxford Levels of Evidence Scale. When possible from the data reported, Cohen’s d-statistic was calculated to evaluate the effect size of reported associations between changes in neck strength and postimpact HN kinematics, concussion risk, and incidence. A Cohen’s d value of 0.2 is considered to be a ‘small’ effect size, 0.5 a ‘medium’ effect size, and ≥ 0.8 a ‘large’ effect size [25]. For the strength training programs, minimum detectable change (MDC95%) values were calculated, when sufficient data was available, to determine the magnitude of change necessary for a resistance training program to produce a clinically meaningful effect on neck strength [26]. | |
|
Table 3: Evidence relating peak isometric strength of the neck musculature and the dynamic stiffness of the HN segment. |
|
Table 4: Model-based evaluation of dynamic stiffness of the HN segment and post-impact kinematics of the head. |
|
Table 5: Modifying effects of anticipatory isometric neck force on dynamic stiffness of the HN segment. |
|
Table 6: Neck strengthening studies in non-clinical populations. |
Results |
|
| Evidence relating neck strength to concussion incidence and risk | |
| Peak isometric strength does not attenuate post-impact kinematics of the head or lower the impact severity of hits to the head, variables commonly used as proxy measures for concussion risk. However, total isometric strength of the neck was found to be a significant predictor of concussion incidence in high school athletes. This level 1b evidence is summarized in Table 3. | |
| The specific association between peak isometric neck strength and concussion incidence has been evaluated in one prospective study [14]. As part of a surveillance study of concussive injuries in three high school sports (basketball, soccer and lacrosse), Collins et al. [14] obtained pre-season measures of peak isometric neck strength for 6,662 high school athletes. Total neck strength was calculated as the mean of the peak isometric force (lbs.) measured in flexion-extension and bilateral side flexion. Concussion incidence was monitored prospectively during the academic years of 2010 and 2011; a clear criterion for concussion diagnosis was not provided. Of the study group, 179 athletes sustained a concussion, which is an incidence rate of 2.7%. Sex- and sport-specific effects were identified. The incidence of concussion was higher in females (P<0.001) and in soccer, where the incidence rate was 5.2 per 10,000 athletes exposures compared to 3.7 in lacrosse and 2.3 in basketball. After adjusting the logistic regression model for sex- and sport-effects, total neck strength remained a significant predictor of concussion incidence (P=0.004). The odds of sustaining a concussion were predicted to decrease by 5% for every one lb. increase in total neck strength. The effect size of total neck strength on concussion incidence was small (Cohen's d=0.29). | |
| In a smaller prospective study [27], higher peak isometric neck strength did not lower the impact severity of hits to the head in minor hockey players. Peak isometric strength of the anterior and anterolateral neck flexors, posterolateral neck extensors and cervical rotators muscle groups was measured prior to the start of the season for thirty-seven elite minor ice hockey players. Participants’ hockey helmets were instrumented with the Head Impact Telemetry (HIT) system to record peak linear and angular acceleration of the head during on-ice head contacts. Head impacts were monitored over 98 games and 99 practices. Post-impact head acceleration profiles were combined with data on the location and duration of impact to yield the Head Impact Telemetry severity profile (HITsp). The HITsp score was used as a criterion of concussion risk in the statistical analysis. Higher peak isometric strength did not predict lower HITsp scores (P≥0.22). | |
| Schmidt et al. [28] confirmed the findings of Mihalik et al. in their prospective study of concussion risk in forty-nine high school and collegiate football players, where again, the criterion for concussion risk was the HITsp score. Peak isometric strength was measured in flexion, extension and bilateral side flexion, with peak magnitudes summed to provide a composite strength score. Football helmets were instrumented with the HIT system and impact kinematics of the head recorded over one season, including both games and practices. HITsp scores were calculated for a total of 19,775 impacts. HITsp scores were rank ordered and the group median used as a cutoff to classify athletes into a ‘high’ or ‘low’ head impact group. HITsp scores were categorized as mild (HITsp<11.7, n=4775), moderate (11.7<HITsp<15.7, n=7309) or severe (HITsp>15.7, n=7691), and logistic regression analysis used to relate HITsp scores to composite strength scores. Higher isometric strength scores did not modify the odds of sustaining a moderate or severe head impact, with an odds ratio of 1.02 (CI95%, 0.80 to 1.32) for moderate impacts and 0.96 (CI95%, 0.67 to 1.36) for severe impacts. By including player position as a covariate in the regression model, Schmidt et al. reported the odds of sustaining a moderate or severe head impact to be highest for linesmen, with an odds ratio of 1.78 (CI95%, 1.01 to 3.16) for moderate impacts and 1.34 (CI95%, 0.29 to 6.23) for severe impacts, despite linesmen having the highest measures of peak isometric neck strength. | |
| While peak isometric neck strength did not predict the odds of sustaining a moderate or severe head impact in prospective sportspecific cohort studies, controlled lab-based studies have described an attenuating effect of peak isometric neck strength on the kinematic response of the HN segment to standardized applications of external forces to the head. These attenuating effects were evaluated using between-subject [11,13,29,30] and within-subject [11,31] experimental designs. This level 2b and 4 evidence is also summarized in Table 3. | |
| Gutierrez et al. [30] correlated peak isometric neck strength, measured in flexion, extension and bilateral side flexion, to postimpact kinematics of the head during controlled soccer ball heading maneuvers in 17 female high school varsity soccer players. They reported a negative correlation between peak measures of isometric neck strength and peak magnitude of post-impact linear acceleration of the head (Pearson’s r, -0.5 to -0.75). While this attenuating effect was significant (P≤0.04), peak isometric strength explained only between 25% and 56% of the variance of post-impact linear acceleration of the head (level 4 evidence). | |
| In contrast to the semi-constrained movement used by Gutierrez et al., other studies have used a pulley system to standardize the application of an external force to the head, either along the sagittal (flexion-extension) plane of HN motion [11,29], or along all three planes of motion of the HN segment [13]. Effects of applied forces were compared between male and female adults, and in athletes, both male and female, 8 to 30 years old, with the a priori assumption that measured differences in post-impact HN kinematics would result from the lower neck strength in females, as well as in children and adolescent athletes. As predicted, adult females exhibited 29% to 49% lower peak isometric strength than adult males and 18% to 29% higher peak post-impact angular acceleration of the head [11,13,29]. From their kinematic data, Mansell et al. [11] and Tierney et al. [29] calculated a 29% lowering in the resistive capacity (or stiffness) of the HN segment in females (level 2b evidence). Additionally, Eckner et al. [13] reported a significant independent effect of age on the resistive capacity of the neck (P<0.001). Peak isometric strength was 32% to 53% lower in athletes of high school age or younger compared to adults, and was associated with 40% higher peak post-impact angular velocity of the head for males and 48% for females (level 2b evidence). | |
| From their data set, Eckner et al. [13] predicted a linear relationship between peak isometric strength and the resistive capacity of the HN segment along the sagittal plane of motion (P<0.02, level 2b evidence), with peak isometric strength explaining 17% to 36% of the variance in post-impact linear and angular velocity of the head (Pearson’s r, 0.42 to 0.60). Peak isometric strength did not predict resistive capacity of the HN segment along the frontal plane or for axial rotation. | |
| Evidence from model-based studies | |
| Model-based simulation provides a method to systematically investigate the specific association between the resistive capacity of the HN segment and post-impact kinematics of the head under different scenarios of external force application. This level 5 evidence is summarized in Table 4. | |
| Using a physical model (Hybrid III dummy), Viano et al. [12] measured the effects of varying the resistive capacity of the HN segment also known as the stiffness, on the post-impact kinematics of the head. The physical force inputs applied to the head component of the Hybrid III model were the mean three-dimensional components of the direction and velocity of external forces recorded by video for 31 head impacts in 25 players of the National Football League who sustained a concussion resulting from helmet-to-helmet or helmetto- ground collisions. Increasing the pre-impact stiffness of the neck component of the Hybrid III model from 80 N/mm, the estimated baseline HN stiffness for the 50th percentile male, to 180 N/mm yielded a 14% attenuation of the peak post-impact linear acceleration of the head, with a 35% lowering of the Head Injury Criterion (HIC). The HIC, calculated as the change in acceleration of the head over the time of force application, is a measure of the likelihood of head injury arising from an impact [32]. The upper limit of 180 N/mm of neck stiffness used in this simulation exceeds the predicted stiffness for the 95th percentile male [12]. The relationship between neck stiffness and post-impact linear acceleration of the head was best described by an exponential function, with relatively small changes in stiffness yielding significant attenuation effects of post-impact head kinematics for lower baseline levels of neck stiffness, with only minor effects for higher baseline levels of stiffness. As an example, a 10N/mm increase from a baseline neck stiffness of 30 N/mm produced a 23% lowering of the HIC compared to the 14% lower HIC with a 40 N/mm increase from a 80 N/mm baseline of neck stiffness. | |
| Shewchenko et al. [33] used a computational model (MADYMO, version 6.0.1, Tass International) to characterize the relationship between stiffness of the HN segment and post-impact kinematics of the head for a simulated soccer ball heading maneuver. In contrast to Viano et al.’s [12] method of increasing stiffness uniformly along all directions of motion, Shewchenko et al. [33], manipulated stiffness of the HN segment indirectly and in direction-specific ways by varying the relative levels of activation across sixty-eight pairs of muscle elements included in the neck model. Activation levels were attributed first to the neck flexor muscle group, with levels adjusted to flex the head and neck toward the ball in preparation for impact. The relative activation levels for the extensor muscle group and sternocleidomastoid muscles were then scaled in iterative fashion to match motions of the HN model to realistic pre- and post-impact kinematics obtained from recorded performances of controlled soccer ball heading maneuvers in seven, non-professional, male soccer players, having five to thirteen years of soccer experience [34]. Resultant forces acting on the upper cervical spine were also predicted. Model-based simulations were then used to evaluate effects of increasing pre-impact muscle activity of the neck flexors to 125% and 150% of their predicted maximum activation, adjusting coactivation levels of extensors and sternocleidomastoid accordingly, on the post-impact kinematics of the HN segment. Raising activation levels to 125% yielded a 20% increase in peak angular acceleration of the head by 20%, with an associated 7% increase in Head Impact Power (HIP), where HIP is a composite index of the rate of energy transfer to the head, estimated by combining peak magnitudes of post-impact linear and angular acceleration of the head [34]. The model also predicted an associated 44% increase in peak magnitude of anterior-posterior shear and 63% increase in axial compression forces at C0-C1. Raising the activation level to 150% did not further influence peak angular acceleration of the head and HIP, with values of 48% and 6%, respectively. However, anterior-posterior shear forces and axial compression forces at the upper cervical spine (i.e., C0-C1 level) were predicted to increase to 79% and 119% of baseline, respectively. These model-based simulations provide evidence of the sensitivity of HN stiffness on parameters of pre-impact muscle activation. | |
| Evidence relating short-latency neck strength to HN kinematics | |
| While there is no consistent evidence for a protective effect of higher peak isometric neck strength in lowering the incidence of concussion or in modifying the post-impact kinematics of the HN segment, there is level 1b [28,35], 2b [11,13,29] and 4 [34,36,37] evidence of an attenuating influence of higher short-latency isometric neck muscle tension, developed prior to impact, on the post-impact kinematics of the HN segment. The attenuating effects of shortlatency neck strength have been evaluated by comparing post-impact kinematics of the HN segment to an externally applied force when the time of impact is either ‘anticipated’ or ‘unanticipated’, with the assumption that knowledge of impact allows individuals to increase isometric tension of their neck muscles and brace for the impact. This level 1b, 2b and 4 evidence is summarized in Table 5. | |
| The attenuating effects of anticipatory pre-tensing of neck muscles on the post-impact kinematics of the HN segment during game play are reported in two prospective cohort studies [28,35]. Mihalik et al. [35] reviewed video capture of on-ice collisions in their study on concussion risk in minor hockey players to determine if upcoming impacts were ‘anticipated’ or ‘unanticipated’. For head impacts of moderate intensity, defined as the range between the 50th to 75th percentile of HITsp scores, anticipation of the contact yielded a 17% attenuation of the peak post-impact angular acceleration of the head (P=0.006; Cohen’s d=0.37), with a 2% lowering of the HITsp scores (P=0.03; level 1b evidence). While significant, the effect size of this attenuating effect was small (Cohen’s d=0.27). | |
| In their study group of forty-nine high school and collegiate football players, Schmidt et al. [28], similarly reported a positive attenuating effect of higher anticipatory HN stiffness. In this study, anticipatory HN stiffness was quantified pre-season using the standard methods of Mansell et al. [11], but scaling the magnitude of the applied external force to body weight. Players with higher anticipatory HN stiffness had lower odds of sustaining moderate and severe head impacts over the football season, with odds ratio of 0.77 (CI95%, 0.61–0.96) for moderate impacts and 0.64 (CI95%, 0.46– 0.89) for severe impacts (level 1b evidence). The effect size of higher anticipatory HN stiffness could not be calculated from the data set reported. | |
| Similar positive effects of anticipation of impact on postimpact HN kinematics were reported in soccer heading maneuvers performed at low (6.2 m/s) and high-speed (7.5 m/s) ball impacts [34]. For low speed head impacts, anticipatory pre-tensing of the neck musculature yielded a 2% attenuation in peak linear acceleration of the head (Cohen’s d=2.12) and a 5% attenuation of peak angular acceleration (Cohen’s d=0.34). This attenuation yielded a 25% reduction in HIP score (Cohen’s d=1.20). Anticipatory pre-tensing of the neck musculature had no effect on post-impact HN kinematics for high-speed impacts. Therefore, anticipatory pre-tensing of neck muscles contributes small to large protective effects on concussion risk only for low-speed impacts (range of Cohen’s d, 0.34 to 2.12, level 4 evidence). | |
| The positive effects of anticipatory pre-tensing of neck muscles in attenuating post-impact kinematics of the HN segment is further supported by level 2b and level 4 evidence from lab-based studies [11,13,29,36,37]. Using their standard methods for quantifying HN stiffness along the sagittal plane, Mansell et al. [11] and Tierney et al. [29] reported a 13% to 21% increase in the resistive capacity of the HN segment with anticipatory pre-tensing of the neck (P≤0.05) [29] and an associated 7% to 24% attenuation of peak magnitudes of post-impact angular acceleration of the head (P≤0.001) [11,29]. Eckner et al. [13] confirmed a positive attenuating effect of anticipatory pre-tensing of the neck on post-impact HN kinematics along all three planes of motion (Pearson’s r=0.42 to 0.66, P<0.001). Reported attenuating responses represent small to large effect size of anticipatory pre-tensing, with Cohen’s d values ranging from 0.03 to 0.70. | |
| Evidence of effectiveness of neck strengthening programs | |
| The second aim of our scoping review was to determine the effectiveness of neck strength training programs in increasing not only peak isometric strength of the neck but as well, the anticipatory or short-latency variables of the force-time strength response of the neck. The parameters of training for the twelve strengthening programs identified by our search strategy are summarized in Table 6. Figure 1 compares the mean (CI95%) effect sizes of training on peak isometric neck strength, stratified by training stimulus - isotonic, elastic, isometric, and isokinetic. | |
|
Figure 1: Effect size (Cohen’s d), with corresponding 95% confidence intervals, is shown for the twelve resistance training programs, stratified by training stimulus: isotonic, elastic, isometric and isokinetic. The boundaries of effect size are identified: α – “small” effect (d=0.20); δ – “medium” effect (d=0.50); φ – “large” effect (d=0.8). The data for male cohort extension strength in the study Mansell et al. [11] has been excluded from the analysis due to a large decrease in extensor strength following the training program for which the authors do not provide an explanation. |
| Calculated MDC95% values for each program are reported in Table 7. Cohen’s d and MDC95% values could not be reliably calculated for two strength training programs, due to insufficient detail of outcome measures [38] and small number of participants (n=5) in the control and strength training groups [39]. | |
|
Table 7: Minimum detectable change of neck strengthening studies. |
| In general, resistance training programs, stratified by training stimulus, produced medium to large effect sizes of change in pre- and post-training measures of peak isometric strength, with Cohen’s d value of 0.65 (CI95%, 0.37 to 0.93) for isotonic, 2.10 (CI95%, 0.74 to 3.41) for isometric, 0.48 (CI95%, 0.11 to 0.86) for isokinetic and 0.47 (CI95%, 0.16 to 0.77) for elastic programs. The widths of the CI95% indicate that the effect size of training varied among specific programs, with some programs producing small effects on strength and others large effects. Of the twelve strength programs appraised in our review, only three produced gains in peak isometric strength exceeding the MDC95% threshold for clinical significance in at least 75% of the direction-specific measurements [39-41]. | |
| The specific effects of neck strength training on the kinematics of the HN segment were evaluated in two studies [11,31]. The program designed by Mansell et al. [11] produced a medium training effect size on neck flexor strength in males (Cohen’s d, 0.54) and large effect sizes on neck flexors and extensors strength in females (Cohen’s d, 1.16 to 1.83). In contrast, the program designed by Lisman et al. [31] produced small training effects on neck strength in flexion, extension and bilateral side flexion (Cohen’s d, 0.13 to 0.34). Based on withinsubject comparisons of the kinematics of the HN segment pre- and post- strength training, neither small or large strength gains were effective in increasing the resistive capacity of the HN segment to externally applied forces. The data from Mansell et al. [11] shows their program did yield a 14% to 48% lowering of the ratio of peak angular acceleration of the head along the flexion-extension plane between ‘anticipated’ and ‘unanticipated’ conditions of external force application. Therefore, a component of their program did positively influence the short-latency anticipatory resistive capacity of the HN segment. | |
Quality of the Research Evidence |
|
| The quality of the research evidence relating isometric neck strength to concussion incidence and risk needs to be evaluated to determine its application to clinical practice. The methodological quality rating (MQR) scores from the PEDro scale and NOS are reported in Tables 3, 5 and 6. The MQR scores could be calculated for the strength training studies, and ranged between 6/10 and 8/10. The main methodological limitations of these studies were nonrandomization of participants, with pre-defined allocation to control and experimental groups due to low number of participants, inability to conceal allocation to the experimental group from participants and assessors, and use of within-subject pre-post assessment rather than between-subject randomization. NOS scores for cohort and case studies ranged between 6/9 and 9/9. Common limitations across studies were low representation of the cohort population, lack of non-exposed control and failure to control for potential confounding variables in statistical models. | |
Interpretation of Current Evidence |
|
| The evidence relating neck strength to concussion incidence in contact sports is very limited, with only one prospective study reporting a small positive effect (Cohen’s d, 0.29) of total isometric neck strength in lowering the incidence of concussion in high school athletes [14]. Based on current evidence, inclusion of neck strength training in the risk management for concussion in contact sports cannot be judiciously recommended. | |
| Research evaluating the effects of neck strength training for concussion risk management is limited in both amount and generalizability of findings. Current evidence from prospective studies is limited by the specific sex and age characteristics of the study groups, with all three studies conducted with adolescent and high-school athletes in whom neuromuscular coordination, physiological cross sectional muscle area, and anthropometric ratio of head-to-neck circumference may be markedly different than in adults [14,27,28]. In a similar way, generalization of evidence relating neck strength to the kinematics of the HN segment under controlled lab-based conditions is inherently limited by reliance on comparative analysis of neck strength between adult females and males [11,29] or adults and youth athletes [13]. As an example, Mansell et al. [11] reported a 29% lowering of the resistive capacity of the HN segment in adult females compared to males and used this between-group difference to infer attenuating effects of higher neck strength on the peak magnitudes of the kinematic response of head to an externally applied force. These measured differences, however, reflect the contribution of factors other than strength, including sex-specific differences in neuromuscular coordination and natural mechanics of the HN segment, as for example higher head-to-neck ratio in females [11,13,29,33,42]. Within-subject designs, using resistance training to manipulate neck strength, should be adopted as a standard to investigate the effects of neck strength on concussion risk. The outcomes of the study by Mansell et al. [11] underline the importance of within-subject designs. In this study, while between-subject differences in strength were related to differences in post-impact HN segments, a relationship between higher neck strength, resulting from resistance training, and peak magnitudes of post-impact HN kinematics could not be defined using within-subject analysis. Therefore, there is a need for multi-centered trials to evaluate the association between neck strength and concussion risk and incidence using within-subject designs in athletic populations of males and females, both youth and adults, and across different contact sports. | |
| Standards for measurement and analysis of neck strength should be adopted to allow for systematic comparison of outcomes across studies. In the eight studies appraised in our review, in which peak isometric neck strength was included as an independent variable in the analysis of concussion incidence and kinematics of the HN segment, strength measures were obtained using a variety of methods: hand-held dynamometry [11,27,29,30]; tensile scale [14]; and custom or commercial fixed-frame dynamometry [13,28,31]. Absence of information regarding the sensitivity and reproducibility of strength measures can lead to errors in interpretation of the outcomes. As an example, from the strength data reported by Collins et al. [14], the mean difference in total strength between the concussed and nonconcussed athletes was calculated to be 1.7 lbs. The researchers used a custom-designed tensile scale to quantify peak isometric neck strength. The tension scale measurements were reported to correlate well with a hand-held dynamometer (Pearson’s r=0.83−0.94, P<0.05) and demonstrated high inter-tester reliability between five different assessors. However, without providing information on the sensitivity of their measurement method, it is not possible to determine if the mean difference of 1.7 lbs. lies outside the 95% confidence interval of the instrument’s measurement error and, therefore, if it is a clinically meaningful difference in strength. As a minimum, researchers need to systematically include MDC95% cutoffs to allow their research findings to be meaningfully evaluated for practice. The positioning of participants for strength measures also varied across assessment protocols, with participants seated and fully restrained below the neck [11,13,29], restrained at the pelvis [31], unrestrained in sitting [14], and restrained and or unrestrained in prone and supine [27,28]. Differences in test position would influence the contribution of other muscles to measured force variables of the neck, again possibly leading to errors in interpretation of outcomes. | |
| Adopting standards for identifying concussion incidence and risk is also recommended. If concussion incidence is used as a primary outcome to evaluate effects of neck strength in contact sports, as in Collins et al.’s study, researchers should adhere to current guidelines and provide a clear statement of assessment tools used [43]. In a similar way, a standard set of kinematic variables should be used to calculate concussion risk which is commonly used as a primary outcome. In reviewed studies, kinematic variables used for the calculation of concussion risk have included peak magnitudes of linear and angular velocity and acceleration of the head, location and duration of impact, as well as combinations of these variables to calculate composite indices of head impact severity including the HITsp, HIC, and the HIP [12,27,28,34,35]. Calculation of concussion risk across these studies appears to be driven by availability of sensor technology rather than by the validity of measures. Only one study was identified in which measured variables of the post-impact kinematics of the head were specifically evaluated in relation to concussion incidence [44]. Broglio et al. [44] used the HIT system to monitor head impacts for seventy-eight high school football players over one season. In total, 54,247 head impacts were analyzed, thirteen of which resulted in a concussion. Using mixed design regression modeling, the set of kinematic variables with the highest predictive value was identified to be the combination of peak angular acceleration of the head along the plane of axial rotation, peak linear acceleration of the head, and location of impact to the front, top, or back of the head. Research is needed to confirm the most appropriate set of kinematic variables predictive of concussions and consistent use of composite scores that incorporate this set of variables, allowing for comparison across studies. As well, as most concussions in contact sports result from forces transmitted to the HN segment by a hit to the body, novel systems may need to be developed to reliably capture HN position and motion without contact information, as well as to lower the price to improve accessibility to systems to support increased use of monitoring systems necessary to develop a large database for multicenter research. | |
| Of the twelve resistance strength training programs critically appraised for their effectiveness in promoting increases in peak isometric strength, only three applied the guidelines of the American College of Sports Medicine (ACSM) for frequency, intensity, time and type (i.e., the F.I.T.T. parameters) [45]. Effective parameters of these programs included: training three to four times per week at a loading intensity of 75% one repetition-maximum (1 RM) or 80% of maximal isometric strength; and increasing the intensity when participants could complete one or more dynamic repetitions beyond the target number [39,40] or when there was an increase in maximal isometric strength [41]. The training intensities used by Mansell et al. [11] and Lisman et al. [31] were conservative by ACSM guidelines, with training intensities of 41% to 53% 1 RM and 45% to 60% 1 RM, respectively. In addition, these two training programs included two training sessions per week and were four weeks shorter than the programs of Taylor et al. [39] and Conley et al. [40]. The conservative intensity of these protocols may not have maximized strength gains which could explain, in part, the absence of an effect of strength training on the resistive capacity of the HN segment. As a standard, MDC95% values should be calculated to ascertain that reported increases in strength with resistance training exceed the probability of error in measurement. It may be that specific MDC95% cutoffs should be established that would allow only meaningful gains in neck muscle strength to be evaluated in terms of their potential benefits in lowering the odds for concussion in contact sports. | |
| The ecological validity of using peak isometric strength as the strength variable of interest in studies of concussion risk management must be considered. Korhonen et al. [46] reported that it takes ≥ 400 ms to reach peak isometric force in skeletal muscles of the lower extremities. Even with the shorter latencies predicted for neck muscles [47], it is unlikely that athletes would have sufficient time to develop their maximum isometric strength in the short-latency required to attenuate post-impact kinematics of the HN segment. However, Almosnino et al. [47] did report that male athletes could develop 50% of their maximal isometric neck strength in 135 to 148 ms. Therefore, facilitating the development of short-latency neck strength should be a primary outcome of neck strength training programs for the risk management of concussion. This recommendation is supported by level 1b, 2b, 3b, and 4 evidence of small to large effects of short-latency anticipatory neck in attenuating magnitudes of post-impact HN kinematics and lowering the severity of head impacts [11,13,28,29,34-37]. | |
| Facilitation of the short-latency rate of isometric force development (RFD) was not specifically addressed in any of the strength training programs appraised. However, several studies have reported significant gains in RFD with resistance training for skeletal muscles of the limbs [15,48-51]. RFD is a velocity-dependent variable of muscle strength that reflects the central activation drive and the mechanics of muscle contraction [15,48]. Therefore, training programs that emphasize high-velocity muscle contractions (i.e., explosive contractions for plyometric movements) have been shown to be effective in facilitating short-latency neuromuscular adaptations to enhance RFD [52,53]. These high-velocity contractions are characterized by high motor neuron firing rates, high muscle force production, and brief contraction times [15,49,54] which increase the absolute magnitude of muscle tension developed in the early phases of a muscle contraction [15,55,56]. Of importance with regards to neck strengthening is that actual mechanical shortening of the muscle is not necessary to elicit short-latency neuromuscular adaptations of RFD; rather, it is the ‘intention’ to produce a high-velocity (or ballistic) contraction that is the effective stimulus [54,55]. Therefore, isometric contractions performed with ballistic intent would be a safe and appropriate strategy to rapidly increase anticipatory early-phase isometric neck muscle strength along all planes of motion, including axial rotation to increase short-latency anticipatory HN stiffness. The direct effects of training with ballistic intent contractions on shortlatency strength and muscle activation was evaluated through a 14- week, high-intensity training strength program of the knee extensor muscle group. The training stimulus used was 4 to 5 sets of heavyto- moderate training loads that ranged from 3 to 10 repetitions maximum [15]. This training program yielded a 17% increase in peak isometric knee extension torque (P<0.001) and a 26%, 22% and 17% increase in RFD at time intervals of 0 to 30 ms, 0 to 50 ms, and 0 to 100 ms, respectively (P<0.05). There was also an increase in the mean level of activation of the quadriceps muscle group by 22% to 143% (P<0.05) from 0 to 100 ms of force onset, and an increase of 41% and 106% from 0 to 75 ms (P<0.01). | |
| The effects of strength training programs on RFD have yet to be systematically investigated within the context of concussion incidence and risk. In our scoping review, only two studies used measures of RFD in their analysis of post-impact kinematics of the HN segment [13,28]. In these studies, RFD was expressed as the maximum slope of the force-time curve to peak muscle force and was reported to be positively associated to increased resistive capacity of the HN segment to controlled applications of external forces to the head [13] and to a lowering of the odds of sustaining head impacts of moderate severity during contact events in games [28]. A standard should be adapted to report RFD measures for discrete time intervals of short-latency force development (e.g., 0 to 25 ms, 0 to 50 ms, and 0 to 100 ms). Almosnino et al. [47,57] demonstrated that short-latency variables of static neck muscle strength could be reliably quantified using standardized methods. With reliable measurement, the theorized protective effects of RFD for concussion could be systematically evaluated. | |
Recommendations for Practice |
|
| Current evidence does not support a benefit of resistance training to increase peak isometric strength as a component of risk management for concussion in contact sports. There is evidence of sufficient level and quality to support further research to specifically evaluate the effects of RFD. At a minimum, RFD should be considered in the evaluation of readiness for return-to-play. If resistance training of the neck is used as a component of athlete preparation, programs should integrate ballistic intent contractions within a motor learning program that will facilitate recruitment of those muscles that are optimally aligned to resist impact. To optimize outcomes, this resistance training approach could be integrated into existing isometric resistance programs with demonstrated effectiveness in producing clinically meaningful changes in peak muscle strength, as for example, the program by Portero et al. [41]. This eight-week isometric strength training program in lateral side flexion produced a 35% increase in peak static strength in seven adult males, 24 to 30 years old, representing a large effect size of (Cohen’s d value, 2.10, CI95%, 0.74 to 3.41). Any program of resistance training used should adhere to ACSM guidelines. | |
| Outcomes of the modeling study by Shewchenko et al. [33] should be considered as a precaution in the design of resistance training programs. Of specific importance are the predicted effects of increasing the level of muscle activation on the anterior-posterior shear and axial compression forces exerted on C0-C1. There is value and need for continued research using computational modeling methods to systematically evaluate effects of modifying peak strength, RFD, HN stiffness, and neuromuscular control on the forces and stability of the cervical spine as per Shewchenko et al.’s [33] approach. | |
| The importance of adopting standardized methods for the assessment and reporting of variables of neck strength cannot be overlooked. As well, affordable methods need to be developed to enhance our general capacity to instrument helmets to monitor head impacts in contact sports. Combining standardized assessment with monitoring into accessible databases would facilitate experimental and computational research of this important topic in concussion risk management. | |
| Most important, any program should emphasis ‘sport-readiness’. Sport intelligence and skill are principal factors that directly influence an athlete’s ability to avoid vulnerable positions or high-risk plays, and to anticipate and prepare for an upcoming impact [35]. Education is also important as athletes need to have knowledge of the specific risks associated with sport participation. This was clearly underscored by the findings of Schmidt et al. [28] that linesmen in high school football were at highest risk for sustaining moderate to severe head impacts, even though they had the strongest necks when compared to other player positions. Lastly, player attitude cannot be overlooked. Even the highest level of preparation cannot lower the risk and incidence of concussion for blindside hits to unsuspecting athletes [1]. This issue of fair play and safe participation must be widely promoted in contact sports. Players must be educated on their responsibilities in assuming roles as both an ‘aggressor’, the player delivering the hit, as well as a recipient of hits. Players should be required to develop skill to safely assume both of these roles, including understanding the purpose of body contact as part of the game being played, skill in delivering hits that are both effective and safe, skill in maintaining awareness of risk for body contact and how to safely receive a hit, and education on the importance of reporting injuries immediately, including concussions. | |
Summary of Key Findings |
|
| Based on current evidence, strength training of the neck musculature cannot be recommended as an effective strategy to lower and incidence of concussion in contact sports. However, one prospective study (level 1b evidence) has provided evidence that higher absolute total isometric neck strength is a significant predictor of concussion incidence in contact sports in high-school athletes. | |
| Higher short-latency isometric neck muscle tension, developed prior to impact, can lower magnitudes of post-impact kinematics of the head (level 1b, 2b, and 4 evidence). Therefore, strength-training programs that facilitate increased gains in short-latency rate of isometric force development may be an important component of neck strength training programs to lower the risk for concussion. | |
| Isometric contractions performed with ballistic intent would be an appropriate strategy to increase the short-latency isometric response of the neck. | |
References |
|
|
|
