Boxing and mixed martial arts: preliminary traumatic neuromechanical injury risk analyses from...

J Neurosurg 116:1070–1080, 2012

1070 J Neurosurg / Volume 116 / May 2012

Historically, linear acceleration has been the main parameter by which to measure athletic head in-jury risk; however, combinations of linear and ro-

tational impact dosage have been widely acknowledged as contributors to head injury. Current standard head in-jury metrics—the GSI9 and HIC35—are based on resul-

tant linear head center of gravity acceleration and cannot be used to quantify rotational injury risk. These injury metrics were developed from laboratory head impact studies conducted decades ago and were correlated to se-vere skull fracture and brain injury, but not less-severe brain injuries like concussion. As HIC, GSI, and resultant linear head center of gravity acceleration are the most widely used and only currently mandated injury metrics for athletic head-padding impact testing,1,19,28 little re-search has been conducted on boxing and mixed martial arts (MMA) padding effects on rotational or combined impact dosage injury risk. Therefore, while boxing and MMA padding may be capable of reducing linear impact dosage (for example, GSI, HIC, linear acceleration, and impact force), the effects of padding on rotationally in-

Boxing and mixed martial arts: preliminary traumatic neuromechanical injury risk analyses from laboratory impact dosage data

Laboratory investigationAdAm J. BArtsch, Ph.d.,1,2 EdwArd c. BEnzEl, m.d.,1–3 VincEnt J. miElE, m.d.,3,4 douglAs r. morr, P.E.,5 And VikAs PrAkAsh, Ph.d.2,6

1Spine Research Laboratory and 3Department of Neurological Surgery, Neurological Institute, Cleveland Clinic; 2Cleveland Traumatic Neuromechanics Consortium; 6Department of Mechanical Engineering, Case Western Reserve University, Cleveland; 5Scientific Expert Analysis (SEA), Ltd., Columbus, Ohio; and 4United Hospital Center Neurosurgery & Spine Center, Clarksburg, West Virginia

Object. In spite of ample literature pointing to rotational and combined impact dosage being key contributors to head and neck injury, boxing and mixed martial arts (MMA) padding is still designed to primarily reduce cranium linear acceleration. The objects of this study were to quantify preliminary linear and rotational head impact dosage for selected boxing and MMA padding in response to hook punches; compute theoretical skull, brain, and neck injury risk metrics; and statistically compare the protective effect of various glove and head padding conditions.

Methods. An instrumented Hybrid III 50th percentile anthropomorphic test device (ATD) was struck in 54 pendulum impacts replicating hook punches at low (27–29 J) and high (54–58 J) energy. Five padding combina-tions were examined: unpadded (control), MMA glove–unpadded head, boxing glove–unpadded head, unpadded pendulum–boxing headgear, and boxing glove–boxing headgear. A total of 17 injury risk parameters were measured or calculated.

Results. All padding conditions reduced linear impact dosage. Other parameters significantly decreased, signifi-cantly increased, or were unaffected depending on padding condition. Of real-world conditions (MMA glove–bare head, boxing glove–bare head, and boxing glove–headgear), the boxing glove–headgear condition showed the most meaningful reduction in most of the parameters. In equivalent impacts, the MMA glove–bare head condition induced higher rotational dosage than the boxing glove–bare head condition. Finite element analysis indicated a risk of brain strain injury in spite of significant reduction of linear impact dosage.

Conclusions. In the replicated hook punch impacts, all padding conditions reduced linear but not rotational impact dosage. Head and neck dosage theoretically accumulates fastest in MMA and boxing bouts without use of protective headgear. The boxing glove–headgear condition provided the best overall reduction in impact dosage. More work is needed to develop improved protective padding to minimize linear and rotational impact dosage and develop next-generation standards for head and neck injury risk.(http://thejns.org/doi/abs/10.3171/2011.12.JNS111478)

kEy words • concussion • traumatic brain injury • boxing • mixed martial arts • injury risk • severity index • rotation

Abbreviations used in this paper: ATD = anthropomorphic test device; CSDM05 = cumulative strain damage measure at the 0.05 level; DDM = dilatational damage measure; GAMBIT = Generalized Acceleration Model for Brain Injury Threshold; GSI = Gadd Severity Index; HIC = Head Injury Criterion; HIP = head impact power; MMA = mixed martial arts; RMDM = relative motion damage measure; SIMon = Simulated Injury Monitor; wPCS = weighted principle component score.

J Neurosurg / Volume 116 / May 2012

Boxing and mixed martial arts injury risk

1071

duced head or neck injury risks remain unclear. Finally, there has been a call for an end to the use of the HIC, a linear acceleration–based criterion derived from the GSI, as a head injury risk quantifier.16

Several investigators have studied athletic padding in sports such as American and Australian football, boxing, soccer, rugby, and equestrian events. In these sports, im-pacts have been replicated using humanoid head forms, an-thropomorphic surrogates, volunteers, or cadavers in im-pacts with kinetic energy from 9 J to 354 J.3,10,12–14,17,18,20,26,34,

37,40 With the exception of one American football helmet study39 and one study of jockey and other equestrian hel-mets,10 rotational head impact dosages have not been re-ported. Interestingly, many studies reported variable lin-ear responses, with some padding increasing and other padding decreasing linear kinematic parameters such as GSI, HIC, and head acceleration.3,7,18,20,26,34,37,40 A study of American football helmets demonstrated that linear and rotational impact dosages were no different in newer helmets and a widely used conventional helmet.37 Most importantly, authors of several studies reported that none of the padding models tested provided adequate head in-jury protection when impacts were compared with bare head impacts.12,18,34,39 Many studies have also indicated that padding thickness, padding composition, and con-tact friction play important roles in reducing head and neck injury risk.7,10,12,18,34,40 Because of these varying head impact attenuation characteristics of athletic padding and the lack of data on padding effects on head and neck inju-ry risk, additional investigations regarding the effects of padding on rotational impact dosage and head and neck injury risk are needed.

MethodsWe investigated linear, rotational, and combined im-

pact dosages of selected boxing and MMA padding gear during a series of 54 pendulum impacts at low (27–29 J) and high (54–58 J) energy to the lateral head of an in-strumented Hybrid III 50th percentile ATD (Humanetics Innovative Solutions). The pendulum impacts were engi-neered to replicate severe hook punches. A total of 5 pad-ding combinations were examined: 1) unpadded (control), 2) MMA glove–unpadded head, 3) boxing glove–unpad-ded head, 4) unpadded pendulum–boxing headgear, and 5) boxing glove–boxing headgear. From these tests, a to-tal of 17 established and proposed head and neck injury risk impact dosage parameters were measured or calcu-lated.

A Hybrid III 50th percentile ATD was instrumented with a triaxial linear accelerometer (Model EAS3–250, Measurement Specialties) rigidly mounted at the head center of gravity and triaxial angular velocity sensor (Model ARS-06S, ATA Sensors) mounted adjacent to the accelerometer. A 6-channel upper neck load cell (Model 1716A, Denton ATD) measured forces and moments in 3 directions. Moments measured by the load cell were translated to the occipital condyles by multiplying respec-tive x and y axis shear forces via the 1.778-cm moment arm. A triaxial linear accelerometer was mounted in the ATD chest to ensure minimal thoracic motion during

impact. All data were collected at 5000 Hz, filtered ac-cording to SAE J21130 and sign convention (direction of positive x, y, and z axes) adhered to SAE J1733.29 Angular acceleration was calculated via filtered angular velocity signal differentiation. The ATD head weighed 5.1 kg with instrumentation, the cervical spine weighed 1.6 kg, and the ATD torso and upper extremities weighed 21.2 kg.

A steel sphere was precisely machined to create a 3.6-kg pendulum impactor mass. The mass was chosen to approximate the effective upper-extremity mass of volunteer boxers during similar punch testing.38 A steel eye hook was threaded into the flat machined face, and the impactor was balanced such that the mass moment of inertia was translated only in the vertical direction. The same steel sphere was used in all trials. A 6.4-mm steel braided cable was hung from a free-swinging carabiner secured to a ceiling strut located approximately 5 m di-rectly above the center of gravity of the ATD head. The ATD was secured to a test stand with tie-down straps and the lower extremities were removed at the femur. Inertial responses were further minimized by securing the test stand with 3500 N of sandbags. The pendulum mass was raised into position via nylon fishing line with 220 N rup-ture strength, tied to the eye hook, and routed through a ceiling-mounted pulley aligned with the head center of gravity, normal to the lateral ATD head surface.

In each head impact trial, the pendulum mass was raised to 0.76 m or 1.52 m and the nylon line was cut with a pair of scissors. The swing heights were selected to rec-reate low- and high-energy impacts on par with boxing punches to the lateral ATD head from prior boxing stud-ies (Pincemaille Y, presented at the 33rd Stapp Car Crash Conference, 1989)27,31,36 and in the range of impact dosage proposed to cause head and neck injury.4,6,8,11,15,21–25,33,41

Tuf-Wear training headgear (0.63 kg), an Everlast Pro Style boxing training glove (0.27 kg), and a UFC Official MMA glove (0.18 kg) were tested. The boxing and MMA gloves were firmly affixed to the pendulum via a combi-nation of double-sided tape on the mass and gaffer’s tape on the exterior. The headgear was secured with double-sided tape on the head and by cinching taut the chinstrap and superior-posterior lacing. The 5 impact conditions are described in Table 1 and shown in Figs. 1–5.

A total of 6 trials were conducted for the 5 impact conditions and the 2 impact energy levels resulting in a total of 54 impacts. The boxing glove–boxing headgear condition was omitted in the low-energy condition. The 6 trials at each swing height were performed in an effort to measure the coefficient of variability and determine statistical power for future testing. Unpadded head im-pacts served as the control for each condition. Because all impacts were laterally directed, the resultant values were analyzed and individual x, y, and z axis components were not examined separately.

For each impact condition, impact dosage was quan-tified via 17 relevant dynamic head and neck injury risk parameters that are discussed in depth in the Appendix. The parameters were grouped into linear, rotational, and combined groups, as shown in Table 2. Some of the more complex parameters included empirical injury risk cri-teria that were calculated post hoc, including the HIC,35

A. J. Bartsch et al.


GSI,9 Generalized Acceleration Model for Brain Injury Threshold (GAMBIT) (Newman JA, presented at the In-ternational Conference on the Biomechanics of Impact, 1986), weighted principle component score (wPCS),11 and head impact power (HIP).22 The HIC and GSI are calcu-lated injury risk functions based on time-varying linear acceleration at the head center of gravity. The GAMBIT, wPCS, and HIP are calculated injury risk functions based on time-varying linear acceleration at the head center of gravity and rigid body angular acceleration. The wPCS has additional empirical relationships that consider GSI and HIC with numerical scaling and offset constants.

Furthermore, linear head center of gravity accelera-tion and angular velocity were used as inputs to the Simu-lated Injury Monitor (SIMon) brain injury risk software32 (V3.051, National Transportation Biomechanics Research Center). The SIMon model consists of a 3D skull and brain with rigid skull, dura-CSF layer, falx cerebri, brain matter, and bridging veins.2,32 The model represents the 50th percentile male with a head mass of 4.7 kg—ap-proximating that of the Hybrid III 50th ATD. There are 10,475 nodes and 7852 elements, with 7776 hexagonal and 76 beam elements. At each time step, SIMon calcu-lates 3 theoretical brain injury risk metrics: 1) the cumu-lative strain damage measure (CSDM05), 2) the relative motion damage measure (RMDM), and 3) the dilatational damage measure (DDM). The CSDM05 is the cumula-tive percentage of brain volume experiencing at least 5% stretch over the duration of head impact loading. This 5% stretch value has been correlated with mild diffuse axonal injury (a result of significant angular brain mo-tion) and transient depolarization.2,32 The RMDM results are an indication of the risk of sustaining acute subdural hematoma due to bridging vein rupture.2,32 More specifi-cally, an RMDM value of 0.5 is equated with an 8% risk, an RMDM value of 1.0 is equated with a 50% risk, and an RMDM value of 2.0 is equated with a 98% risk of acute subdural hematoma. A third brain injury metric known as the DDM, which was developed as an estimate of vacuum contusions within brain tissue (Bandak FA, presented at the Stapp Car Crash Conference, 1994),2,32 was also cal-culated for all impacts.

The mean maximum value (mean of the maximum values from each of 6 trials) for each parameter (kine-matic data, kinetic data, and SIMon) in each condition was used in a 2-tailed paired t-test comparison against the unpadded control condition. The null hypothesis was

that the mean maximum values of the variables compared for each condition were equal. The alternative hypothesis was that the mean maximum values were not equal. A 5% significance (a = 0.05) level was used. This hypothesis was tested using a 2-sample Welch t-test, with unequal sample sizes and unequal variances, assuming a Student t-distribution and requiring the Satterthwaite approxima-tion to determine degrees of freedom.

TABLE 1: Impact matrix

ConditionImpact Energy (J)

Impact Momentum (N-sec)

Low High Low High

unpadded (control) 27 54 14 20unpadded pendulum– headgear

27 54 15 20

MMA glove–bare head 28 56 15 21boxing glove–bare head 29 58 14 21boxing glove–headgear — 58 — 21

Fig. 1. Photograph demonstrating the unpadded (control) impact condition.

Fig. 2. Photograph showing the bare pendulum–headgear condi-tion.



1073

ResultsIn response to the hook punch impacts recreated by

the impact pendulum, there was some variability in kine-matic (linear head center of gravity acceleration, angular velocity) and kinetic (neck force, neck moment) indepen-dent variables during the sequence of 6 repeated impacts. As a means of quantifying this variability for the data

collected, we have collated in Table 3 the respective coef-ficient of variation of each impact condition for the set of 6 repeated impact trials. The coefficient of variation was calculated by dividing the standard deviation from each set of 6 impacts by the mean value at the time of high-est linear center of gravity acceleration resultant. In spite of best attempts to ensure consistent impacts, the average coefficient of variation shown in Table 3 varied from as low as 6.5% (unpadded control) to as high as 26.1% (box-ing glove–boxing headgear). The boxing glove–boxing headgear trials demonstrated the highest variation.

The mean resultant values for ATD head center of gravity linear acceleration, head angular velocity, head angular acceleration, and impact force from the impact conditions are presented in Figs. 6–9. A representative example of the computational simulation results of the SIMon software32 used to generate CSDM, DDM, and RMDM brain injury risk measures is presented in Fig. 10.

A total of 17 dynamic head and neck impact dosage parameters were then compared, via a 2-tailed t-test, to the unpadded control impacts for the low- and high-en-ergy conditions. The results are summarized in Tables 4 and 5. (The maximum value for each of the parameters is shown.) Based on this t-test comparison with the unpad-ded control condition, parameters that were significantly lower at the p < 0.05 level are shaded and parameters that were significantly higher are shown in boldface. Because the DDM results reported by SIMon were near zero for all impacts, this parameter was not reported.

Low-Energy ResultsAs seen in Table 4, for the linear impact dosage t-test

results, impact duration increased significantly (p < 0.05) from the unpadded control, whereas all other parameters except linear momentum (including linear acceleration and GSI) significantly decreased across all conditions. The MMA glove–bare head and unpadded pendulum–headgear conditions showed a significant increase in linear momentum transfer. While angular acceleration was significantly reduced in all 3 conditions compared

Fig. 3. Photograph of the MMA glove–bare head condition.

Fig. 4. Photograph showing the boxing glove–bare head condition.

Fig. 5. Photograph demonstrating the boxing glove–headgear condition (high-energy trial only).



with the unpadded control, angular velocity and angular momentum were significantly reduced in the unpadded pendulum–headgear condition only. Angular momentum was significantly increased in the MMA glove–bare head and boxing glove–bare head conditions. The neck mo-ment was significantly reduced in the MMA glove–bare head and unpadded pendulum–headgear conditions, but not in the boxing glove–bare head condition. All of the low-energy combined impact dosage parameters were

significantly decreased except for head kinetic energy in the MMA glove–bare head and boxing glove–bare head conditions, as well as the CSDM05 in the boxing glove–bare head condition.

High-Energy ResultsIn the high-energy tests (Table 5), impact duration

was significantly increased (p < 0.05) across all conditions. The other linear impact dosage parameters, with the ex-ception of linear momentum, were significantly reduced. The linear momentum showed a significant increase in 3 of the 4 high-energy conditions studied. The high-energy rotational impact dosage parameters were associated with greater variation than the low-energy comparisons. Angu-lar velocity significantly increased in MMA glove–bare head and boxing glove–bare head conditions, significantly decreased with unpadded pendulum–headgear, and had no statistically significant difference in boxing glove–headgear conditions. Angular momentum results were the same with the exception of a significant increase in the boxing glove–headgear condition. Angular acceleration significantly decreased in the unpadded pendulum–head-gear and boxing glove–headgear conditions but was no different in the MMA glove–bare head and boxing glove–bare head conditions. Neck moment was significantly re-duced for the MMA glove–bare head and unpadded pen-dulum–headgear conditions but no different for the other 2 conditions. When examining the combined impact dosage parameter results, kinetic energy transferred to the ATD head was significantly higher for both the MMA glove–bare head and unpadded pendulum–headgear conditions. The calculated GAMBIT, wPCS, and HIP were signifi-cantly decreased versus the unpadded control values for all conditions. The CSDM was significantly reduced in the 2 headgear conditions, but significantly larger in the MMA glove–bare head and boxing glove–bare head con-ditions. Finally, the RMDM significantly decreased with the exception of the MMA glove–bare head condition.

TABLE 2: Impact dosage parameters*

Parameter Units

linear parameters linear acceleration g GSI NA impact duration sec linear momentum transfer N-sec neck force N impact force Nrotational parameters angular velocity rad/sec angular acceleration rad/sec2

angular momentum transfer N-m-sec neck moment N-mcombined parameters kinetic energy transfer J GAMBIT NA wPCS NA HIP NA CSDM05 NA RMDM NA DDM NA

* NA = not applicable.

TABLE 3: Coefficient of variation of each resultant independent variable*

ParameterUnpadded (control)

Unpadded Pendu- lum–Headgear

MMA Glove– Bare Head

Boxing Glove– Bare Head

Boxing Glove–Headgear

low energy linear acceleration 4.7 6.9 11.8 12.4 — angular velocity 10.3 24.6 10.0 19.9 — neck force 7.4 21.1 22.0 16.9 — neck moment 15.1 14.0 27.4 16.9 — average 9.4 16.7 17.8 16.5 —high energy linear acceleration 2.1 4.3 13.5 9.9 37.7 angular velocity 12.7 29.0 10.2 9.3 33.3 neck force 3.5 5.9 10.8 9.3 23.3 neck moment 7.8 10.5 20.2 7.8 9.9 average 6.5 12.4 13.7 9.7 26.1

* Values taken at time corresponding to maximum linear acceleration.



1075

DiscussionAs a means to study boxing and MMA padding effects

on head and neck injury risk, we preliminarily investigat-ed 17 dynamic head and neck impact dosage injury risk parameters in a series of 54 low- and high-energy impacts to an instrumented Hybrid III ATD. Our impacts recreated real-world right hook punches to the left side of the head and involved 5 padding combinations: unpadded (control), MMA glove–bare head, boxing glove–bare head, unpad-ded pendulum–headgear, and boxing glove–headgear. Measured linear head acceleration (54.7–144 g), angular acceleration (1740–5550 rad/second2), angular velocity (12.8–26.2 rad/second), neck moment (18.5–47.7 N-m), and impact force (3080–8900 N) dosages were within ranges of theoretical head and neck injury limits.3,5,10,12–14,

17,18,20,26,34,37,40

Hand PaddingAll low- and high-energy padding conditions re-

duced linear-based impact dosage, including linear accel-

eration, GSI, neck force, and impact force. This finding was expected; padding has historically been designed to provide protection against linear dosage. However, when we examined rotational impact dosage results, and spe-cifically for 2 of the real-world conditions (MMA glove–bare head and boxing glove–bare head), we found that some conditions reduced rotational impact dosage better than others. For example, the boxing glove–bare head condition always significantly reduced angular velocity, angular acceleration, angular momentum transfer, and neck moment compared with the unpadded control. In contrast, however, the MMA glove–bare head condition always significantly increased angular velocity and angu-lar momentum transfer while having no effect on neck moment. The increased angular velocity and angular mo-mentum were perhaps related to the fact that the MMA glove made contact for a greater time duration than did the boxing glove (0.0172 vs 0.0136 seconds in low-energy impacts and 0.0143 vs 0.0073 seconds in high-energy im-pacts). Therefore, the MMA glove induced greater angu-lar velocity and momentum to the bare head by virtue of

Fig. 6. Graphs illustrating the linear acceleration in the high-energy (upper) and low-energy (lower) trials. Fig. 7. Graphs showing the angular velocity in the high-energy (up-

per) and low-energy (lower) trials.



this greater duration of contact compared with the boxing glove–bare head condition.

The contact duration differences between the MMA–bare head and boxing glove–bare head scenarios may be due to the differences in padding compression,10,12,18,34,40 coefficient of friction,7 and the different glove masses in the different conditions. The only caveat was that contact duration appeared to have little or no effect on angular ac-celeration, as this parameter was reduced for both MMA glove–bare head and boxing glove–bare head low-energy conditions, and at high energy the boxing glove–headgear condition had the highest contact duration but lowest an-gular acceleration. Because angular acceleration has been routinely cited as a key rotational injury risk metric, our findings point to the need to consider additional rotation-al parameters like angular velocity, angular momentum transfer, and neck moment when quantifying head and neck impact dosage from boxing and MMA padding and impact conditions.

Although some real-world conditions significantly increased (high-energy boxing glove–bare head) or de-

creased (low-energy boxing glove–bare head) kinetic en-ergy transfer, these differences were probably not very meaningful from an engineering or clinical perspective. The calculated injury risk functions GAMBIT, wPCS, and HIP were all significantly reduced for all conditions.

Advancing beyond empirical injury risk functions, the SIMon finite element brain injury model provided more specific insight into brain stretching, compression, and pressure. Hence, SIMon predicted a significantly lower risk of acute subdural hematoma (via the RMDM) and negligible risk of vacuum contusion (via the DDM) for the MMA glove–bare head, boxing glove–bare head, and boxing glove–headgear conditions. However, when comparing risk of diffuse axonal injury defined by a 5% brain strain (via the CSDM05), the SIMon-predicted CDSM05 values for the boxing glove–bare head and box-ing glove–headgear conditions were significantly lower than the value for the unpadded control condition, while the predicted value for the MMA glove–bare head condi-tion did not differ significantly from that of the unpadded control in the low-energy testing and was significantly higher than the control value in the high-energy testing.

Fig. 8. Graphs demonstrating the angular acceleration in the high-energy (upper) and low-energy (lower) trials.

Fig. 9. Graphs illustrating the impact force in the high-energy (up-per) and low-energy (lower) trials.



1077

This was consistent with the increased rotational impact dosage associated with the MMA glove–bare head condi-tion, which in turn was related to both greater rotational kinetic energy transmission and greater contact duration. SIMon generates injury risk results based on combined linear–rotational kinematic inputs. Hence, diffuse axonal

injury (the etiology of which is thought to be, in part, re-lated to rotational forces) may be expected to occur at a higher rate in MMA glove–bare head impacts than in similar boxing glove–bare head impacts. Moreover, it is important to note that the boxing glove–bare head con-dition had the best combined-parameter results, showing the largest reduction of all 3 real-world conditions for GAMBIT, wPCS, HIP, CSDM05, RMDM, and DDM. Fi-nally, although significantly reduced compared with the unpadded control, the RMDM results for all 3 real-world conditions exceeded the theoretical injury threshold of 1.00 and indicated a heightened risk of acute subdural he-matoma for the theoretical impact conditions we studied.

Head PaddingThe existing literature on head padding for athletes

has shown variable linear and rotational head and neck injury risk mitigation, dependent upon impact condi-tions.3,10,18,20,26,34,37,40 In particular, a recent study examin-ing National Football League (NFL) football helmets37 found that of the 4 newer helmet models tested against an existing helmet model, none of the new helmets consis-tently reduced linear or rotational head and spine injury risk during a series of 10 real-world impact recreations. It should be noted that the NFL study conducted descrip-tive statistical analyses and could not be directly evalu-ated with the comparative statistical analysis from this study. Regardless, of the 3 real-world conditions (MMA

TABLE 4: Maximum values from the low-energy trials*

Parameter Unpadded (control)Unpadded Pendulum–

HeadgearMMA Glove–Bare

HeadBoxing Glove–Bare

Head

linear parameters head acceleration (g) 153 54.7 54.7 62.0 GSI 281 51 42 56 impact duration (sec) 0.0037 0.0157 0.0172 0.0136

linear momentum transfer (N-sec) 10.8 11.3 10.6 11.2 neck force (N) 778 397 478 518 impact force (N) 8390 3080 3190 3920 rotational parameters angular velocity (rad/sec) 15.8 16.7 18.5 12.8 angular acceleration (rad/sec2) 4810 2960 2440 2450 angular momentum transfer (N-m-sec) 0.269 0.305 0.324 0.245 neck moment (N-m) 32.7 18.5 34.0 23.5 combined parameters kinetic energy transfer (J) 11.6 12.7 11.0 10.9 GAMBIT 0.624 0.236 0.230 0.257 wPCS 87.4 31.8 29.6 33.2 HIP 9720 3930 3670 3860 CSDM05 0.606 0.397 0.448 0.195 RMDM 1.89 1.43 1.16 1.27 DDM — — — —

* Values are means of the maximum values from 6 impacts for each condition. Boldface indicates significant increase (p < 0.05); dark shading indicates significant decrease (p < 0.05).

Fig. 10. Representative frame from the SIMon computational simu-lation results displaying pressure isosurfaces within the brain from box-ing glove–bare head impacts.



glove–bare head, boxing glove–bare head, and boxing glove–boxing headgear) studied, the boxing glove–box-ing headgear condition had the most meaningful reduc-tion in most of the parameters quantified and should pro-vide the best overall head and neck injury protection.

Study VulnerabilitiesA confounding factor associated with the methodol-

ogy employed here was the variable mass of the impactor (simulated gloved hand) and ATD head. We conducted impacts at the higher range of human punch magnitude, while still being below impact magnitudes found in known concussive collisions.6,36,38 However, due to the changing mass of the impactor and ATD head, impact momentum and kinetic energy varied by as much as 7.5% and 7.8%, respectively. Therefore, the increased mass of the impactor or ATD head may have affected the post-impact head momentum and kinetic energy results. The significance of our results may have been different if impact momentum and/or energy were held constant be-tween conditions. This, however, would have required the alteration of head and glove mass and/or impact velocity, which would have eliminated one source of variability by adding another. A further confounding variable was that the ATD head form’s vinyl “skin” did not exactly re-produce human cranium-skin gliding, contact friction, or sweat. Hence, especially for the boxing glove–bare head and MMA glove–bare head conditions, the impact dos-

age outcomes may be different in a live person due to the differences between the human and ATD head form.

ConclusionsIn our study, which replicated hook punches to the

side of the head, results indicated that all padding condi-tions reduced linear impact dosage, including linear ac-celeration, GSI, neck force, and impact force. Our results suggest that head and neck impact dosages accumulate fastest in MMA and boxing conditions absent protec-tive headgear. Additional injury risk parameters that in-cluded rotational kinetics and kinematics significantly decreased, significantly increased, or were unaffected depending on whether the impactor, head, or both, were padded. The SIMon finite element brain injury model indicated heightened theoretical risk of injurious brain strain injury for boxers or mixed martial artists regard-less of padding used.

Of the 3 real-world conditions (MMA glove–bare head, boxing glove–bare head, and boxing glove–box-ing headgear) studied, the boxing glove–boxing headgear condition had the most meaningful reduction in most of the parameters quantified and should provide the best overall head and neck injury protection for competitors. The MMA glove–bare head condition resulted in linear impact dosage that was superior to the boxing glove–bare head condition (that is, it was associated with a lower in-

TABLE 5: Maximum values from the high-energy trials*

Parameter Unpadded (control)Unpadded Pendu-

lum–HeadgearMMA Glove– Bare Head

Boxing Glove–Bare Head

Boxing Glove–Headgear

linear parameters head acceleration (g) 232 129 117 144 65.0 GSI 768 265 219 311 79 impact duration (sec) 0.0035 0.0114 0.0143 0.0073 0.0182

linear momentum transfer (N-sec) 15.2 15.5 15.3 16.3 15.9 neck force (N) 1,252 746 836 868 629 impact force (N) 12,770 7,160 6,630 8,900 4,240 rotational parameters angular velocity (rad/sec) 18.5 22.6 26.2 16.3 18.3 angular acceleration (rad/sec2) 5,260 5,550 5,240 3,800 1,740 angular momentum transfer (N-m-sec) 0.345 0.406 0.449 0.321 0.364 neck moment (N-m) 48.8 34.0 47.7 38.2 42.4 combined parameters kinetic energy transfer (J) 22.7 23.7 23.1 23.3 22.1 GAMBIT 0.936 0.541 0.491 0.583 0.264 wPCS 173.2 80.9 71.3 87.8 35.8 HIP 20,500 12,700 11,600 12,700 5,440 CSDM05 0.636 0.774 0.819 0.452 0.210 RMDM 2.35 2.48 2.17 2.03 1.08 DDM NA NA NA NA NA

* Values are means of the maximum values from 6 impacts for each condition. Boldface indicates significant increase (p < 0.05); dark shading indicates significant decrease (p < 0.05).



1079

jury risk), but possibly due to increased contact duration with the ATD head, induced rotational impact dosage greater than the boxing glove–bare head condition (that is, it was associated with greater injury risk).

While extrapolating to the real world is difficult, our results point to the need to understand more about how single impact dosage or dosage accumulation over time may cause a heightened risk of acute and subacute injury as a function of the boxing and MMA protective padding used. Hence, the multivariate analysis presented here pro-vides a methodology for studying this dosage accumula-tion in striking sports. Also, our results emphasize the need to examine protective athletic padding to minimize rotational impact dosage accumulation. Finally, this study clearly highlights the need to consider linear and rota-tional head and neck injury risk in developing impact test standards for next-generation protective padding.

Disclosure

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. They acknowledge assistance for the study from the National Institutes of Health Ruth L. Kirchstein T32 Training Grant AR050959, Cleveland Clinic Center for Spine Health, and Ohio Third Frontier. SEA, Ltd. provided use of the testing facility, impactor, instrumentation, data acquisition equipment, and Hybrid III ATD.

Author contributions to the study and manuscript prepara-tion include the following. Conception and design: all authors. Ac quisition of data: Bartsch, Miele, Morr. Analysis and interpreta-tion of data: Bartsch, Morr, Prakash. Drafting the article: Bartsch, Morr, Prakash. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final ver-sion of the manuscript on behalf of all authors: Bartsch. Statistical analysis: Bartsch, Morr, Prakash. Study supervision: all authors.

Appendix

This article contains an appendix that is available only in the online version of the article.

References

1. American Society for Testing and Materials: Standard Test Methods for Equipment and Procedures Used in Evaluat-ing the Performance Characteristics of Protective Head-gear. Designation: F1446-01a. West Conshohocken, PA: ASTM International, 2002

2. Bandak FA, Zhang AX, Tannous RE, DiMasi F, Masiello P, Eppinger R: SIMon: a simulated injury monitor; application to head injury assessment. National Highway Traffic Safety Administration (NHTSA) Website. (http://www-nrd.nhtsa.dot.gov/pdf/esv/esv17/proceed/00222.pdf) [Accessed January 12, 2012]

3. Broglio SP, Ju YY, Broglio MD, Sell TC: The efficacy of soc-cer headgear. J Athl Train 38:220–224, 2003

4. Casson IR, Pellman EJ, Viano DC: Concussion in the National Football League: an overview for neurologists. Phys Med Re-habil Clin N Am 20:195–214, x, 2009

5. Caswell SV, Deivert RG: Lacrosse helmet designs and the ef-fects of impact forces. J Athl Train 37:164–171, 2002

6. Duma SM, Manoogian SJ, Bussone WR, Brolinson PG, Go-forth MW, Donnenwerth JJ, et al: Analysis of real-time head accelerations in collegiate football players. Clin J Sport Med 15:3–8, 2005

7. Finan JD, Nightingale RW, Myers BS: The influence of re-duced friction on head injury metrics in helmeted head im-pacts. Traffic Inj Prev 9:483–488, 2008

8. Fréchède B, McIntosh AS: Numerical reconstruction of real-life concussive football impacts. Med Sci Sports Exerc 41: 390–398, 2009

9. Gadd CW: Use of a weighted-impulse criterion for estimating injury hazard, in Proceedings of the 10th Stapp Car Crash Conference. New York: Society of Automotive Engineers, 1966, Vol 10, pp 164–174

10. Gilchrist A, Mills NJ: Protection of the side of the head. Accid Anal Prev 28:525–535, 1996

11. Greenwald RM, Gwin JT, Chu JJ, Crisco JJ: Head impact se-verity measures for evaluating mild traumatic brain injury risk exposure. Neurosurgery 62:789–798, 2008

12. Hrysomallis C: Impact energy attenuation of protective foot-ball headgear against a yielding surface. J Sci Med Sport 7: 156–164, 2004

13. Knouse CL, Gould TE, Caswell SV, Deivert RG: Efficacy of rugby headgear in attenuating repetitive linear impact forces. J Athl Train 38:330–335, 2003

14. Lewis LM, Naunheim R, Standeven J, Lauryssen C, Richter C, Jeffords B: Do football helmets reduce acceleration of impact in blunt head injuries? Acad Emerg Med 8:604–609, 2001

15. Margulies SS, Thibault LE: A proposed tolerance criterion for diffuse axonal injury in man. J Biomech 25:917–923, 1992

16. McElhaney JH: In search of head injury criteria. Stapp Car Crash J 49:v–xvi, 2005

17. McIntosh A, McCrory P, Finch CF: Performance enhanced headgear: a scientific approach to the development of protec-tive headgear. Br J Sports Med 38:46–49, 2004

18. McIntosh AS, McCrory P: Impact energy attenuation perfor-mance of football headgear. Br J Sports Med 34:337–341, 2000

19. National Operating Committee on Standards for Athletic Equipment (NOCSAE): Standard test method and equipment used in evaluating the performance characteristics of protec-tive headgear/equipment. NOCSAE Website. (http://www.nocsae.org/standards/pdfs/Standards%20’10/ND001-08m10-Drop%20Impact%20Test%20Method%20.pdf) [Accessed January 12, 2012]

20. Naunheim RS, Ryden A, Standeven J, Genin G, Lewis L, Thompson P, et al: Does soccer headgear attenuate the impact when heading a soccer ball? Acad Emerg Med 10:85–90, 2003

21. Newman JA, Beusenberg MC, Shewchenko N, Withnall C, Fournier E: Verification of biomechanical methods employed in a comprehensive study of mild traumatic brain injury and the effectiveness of American football helmets. J Biomech 38:1469–1481, 2005

22. Newman JA, Shewchenko N, Welbourne E: A proposed new bio-mechanical head injury assessment function—the maximum power index. Snell Memorial Foundation Website. (http://www.smf.org/docs/articles/hic/Newman_Max_Power_ Index.pdf) [Accessed January 12, 2012]

23. Ommaya AK, Goldsmith W, Thibault L: Biomechanics and neuropathology of adult and paediatric head injury. Br J Neu-rosurg 16:220–242, 2002

24. Ommaya AK, Hirsch AE: Tolerances for cerebral concus-sion from head impact and whiplash in primates. J Biomech 4:13–21, 1971

25. Pellman EJ, Viano DC, Tucker AM, Casson IR, Waeckerle JF: Concussion in professional football: reconstruction of game impacts and injuries. Neurosurgery 53:799–814, 2003

26. Smith PK, Hamill J: The effect of punching glove type and skill level on momentum-transfer. J Hum Mov Stud 12:153–161, 1986

27. Smith TA, Bishop PJ, Wells RP: Three dimensional analysis of linear and angular accelerations of the head experienced



in boxing, in Proceedings of the International Research Council on Biomechanics of Impact, Vol 1. Bergish-Glad-bach, Germany: IRCOBI, 1973–1989, pp 271–286

28. Snell Memorial Foundation: Standard for protective headgear for use in non-motorized sports. Snell Memorial Foundation Website. (http://www.smf.org/standards/n94/n94std) [Ac-cessed January 12, 2012]

29. Society of Automotive Engineers: SAE Information Report. Sign Convention for Vehicle Crash Testing—SAE J1733. Warrendale, PA: Society of Automotive Engineers, 1994

30. Society of Automotive Engineers: SAE Recommended Prac-tice. Instrumentation for Impact Test—Part 1—Electronic Instrumentation—SAE J211/1. Warrendale, PA: Society of Automotive Engineers, 1995

31. Stojsih S, Boitano M, Wilhelm M, Bir C: A prospective study of punch biomechanics and cognitive function for amateur boxers. Br J Sports Med 44:725–730, 2010

32. Takhounts EG, Eppinger RH, Campbell JQ, Tannous RE, Power ED, Shook LS: On the development of the SIMon finite element head model. Stapp Car Crash J 47:107–133, 2003

33. Thibault LE, Gennarelli TA: Brain injury: an analysis of neu-ral and neurovascular trauma in the non-human primate. Ann Proc Assoc Adv Automot Med 34:337–351, 1990

34. Tierney RT, Higgins M, Caswell SV, Brady J, McHardy K, Driban JB, et al: Sex differences in head acceleration during heading while wearing soccer headgear. J Athl Train 43: 578–584, 2008

35. Versace J: A review of the severity index, in Proceedings of the 15th Stapp Car Crash Conference. New York: Society of Automotive Engineers, 1971, pp 771–796

36. Viano DC, Casson IR, Pellman EJ, Bir CA, Zhang L, Sher-man DC, et al: Concussion in professional football: compari-son with boxing head impacts—part 10. Neurosurgery 57: 1154–1172, 2005

37. Viano DC, Pellman EJ, Withnall C, Shewchenko N: Concus-sion in professional football: performance of newer helmets in reconstructed game impacts—part 13. Neurosurgery 59: 591–606, 2006

38. Walilko TJ, Viano DC, Bir CA: Biomechanics of the head for Olympic boxer punches to the face. Br J Sports Med 39:710–719, 2005

39. Withnall C, Shewchenko N, Gittens R, Dvorak J: Biomechani-cal investigation of head impacts in football. Br J Sports Med 39 (Suppl 1):i49–i57, 2005

40. Withnall C, Shewchenko N, Wonnacott M, Dvorak J: Effec-tiveness of headgear in football. Br J Sports Med 39 (Suppl 1):i40–i48, 2005

41. Zhang L, Yang KH, King AI: A proposed injury threshold for mild traumatic brain injury. J Biomech Eng 126:226–236, 2004

Manuscript submitted August 23, 2011.Accepted December 12, 2011.Please include this information when citing this paper: published

online February 7, 2012; DOI: 10.3171/2011.12.JNS111478.Address correspondence to: Adam J. Bartsch, Ph.D., Cleveland

Clinic Spine Research Laboratory, Luth2-C, 1730 West 25th Street, Cleveland, Ohio 44113. email: [email protected].

Boxing and mixed martial arts: preliminary traumatic neuromechanical injury risk analyses from...

Documents

Transcript of Boxing and mixed martial arts: preliminary traumatic neuromechanical injury risk analyses from...