Posture and Helmet Configuration Effects on Joint Reaction Loads in the Middle Cervical Spine

METHODS

Materials

We used a 24-degree of freedom cervical spine model in OpenSim 4.1 ( Fig. 1 ). It is equipped with passive elements representing the intervertebral discs and ligaments, and 216 active muscle elements representing 29 muscles bilaterally in the cervical spine. ⁶ It was designed to map the total head-trunk angle along with collected EMG channels for compression and shear forces at each vertebral joint in the cervical spine. The overall head-trunk angle was partitioned among the eight intervertebral joints (including the skull-C1 and C7-T12) in proportion to their range of motion. The linear enveloped EMG was used to drive a Hill-type muscle model to estimate the force in each muscle fascicle. An EMG-assisted optimization routine then adjusted each muscle force to ensure that there was agreement between the EMG-derived moment estimations and the net joint moments calculated from inverse dynamics. ^8,14

View Figure

• Download as Powerpoint
• Download Figure

The helmet (Gentex HGU-56/P; Gentex, Carbondale, PA, USA) was represented as a separate segment in the neck model, securely fastened to the skull segment with a weld joint. For the normal helmet condition, the center of mass of the helmet was assumed to be coincident with the center of mass of the skull. The center of mass of the other helmet conditions were calculated from CAD drawings and represented as offsets from the standard helmet center of mass locations ( Table I ).

View Fullscreen

Procedure

Following a brief warm up, subjects were fixed to a chair with Velcro straps to isolate cervical spine movement while wearing a Canadian Armed Forces CH146 Griffon helicopter helmet. The helmet was mounted on a robotic arm (HP50 with NX100 Controller, Yakasawa Motoman Robotics, Mississauga, Ontario, Canada). Maximum voluntary exertions (MVE) were performed isometrically against the resistance of the robot in three directions: pressing forward (flexion), backward (extension), or to the right (lateral bending). The protocol is illustrated in Fig. 1A, but the interested reader is directed to McKinnon et al. ²⁹ for additional details.

Kinematic data from a passive motion capture system (Vicon MX, Vicon Motion Systems Ltd., Los Angeles, CA, USA) were sampled at 50 Hz from a total of nine markers. Markers from the xiphoid process, acromion process, and C7 spinous process were used to make the local coordinate system (LCS) for the trunk, while the markers over the ears and the most superior point on the helmet were used to make the LCS for the head/helmet system. The relative rotation between these two LCSs was used to quantify the overall head-trunk angle used as an input into the cervical spine model. ⁶ A self-selected neutral posture was obtained by instructing participants to “look straight ahead” during a seated calibration trial.

Following MVEs, the helmet was decoupled from the robotic arm and the experimental trials were completed. Four helmet conditions were tested (Fig. 1B): no helmet (No Helmet), helmet only (hOnly), helmet with the NVGs deployed (hNVG), and helmet with the counterweight and NVG (hCW + NVG). Starting from a neutral posture, subjects moved to one of five target postures, which they held for 15 s before returning to the neutral posture. The target postures were: 45° of flexion (45flex), 30° of extension (30ext), 45° of axial rotation (to the left), 20° of lateral bending (to the left), and a neutral posture (no movement) (Fig. 1B). Real time Euler-Cardan (XYZ sequence) angles were calculated in accordance to ISB standards, ⁴² and shown to the participant for kinematic feedback on their posture. These postures were selected as they tend to be some common ones undertaken by flight crew. ¹³

Pregelled (2 cm interelectrode distance) Ag/AgCl surface EMG electrodes recorded the muscular activity of 10 muscles (5 bilaterally): splenius capitis, sternocleidomastoid, levator scapulae, the splenii and semispinalis (cervical erectors), and the upper trapezius. Briefly, the upper trapezius was palpated against light resistance to shrugging and the sternocleidomastoid to axial rotation. These two muscles form the borders to the neck’s posterior triangle situated on the lateral aspect of the neck, where the superficial portions of splenius capitis and levator scapulae are accessible for surface EMG. Finally, cervical erector electrodes were placed roughly 2 cm on either side of the midline at approximately the C2 level. The interested reader is referred to more detailed documentation on this protocol. ^23,33 A 16-channel Noraxon Telemyo 2400T G2 Telemetry electromyography system (Noraxon USA Inc., Scottsdale, AZ, USA) amplified the signals which were sampled at 1500 Hz through a 16-bit analog to digital card. The resulting signal was linear-enveloped by de-trending, full-wave rectifying, and low-pass filtering with a single pass low-pass critically damped filter with cutoff frequencies chosen to mimic the electromechanical delay of the underlying muscle, which were adapted from Almosnino et al. ² ( Table II ). Muscles with no documented electromechanical delay were assigned a cutoff frequency of 4 Hz, as this has been used in previous cervical spine investigation. ²⁵ Each muscle was normalized to the maximum from the MVE trials.

View Fullscreen

RESULTS

There was a significant interaction effect between posture and helmet configuration (Wald χ²: 25.388; df = 12; P = 0.01309). Deviated postures tended to increase compression, but the effect was not consistent across helmet conditions, indicating a nonlinear relationship between added head mass and joint compression ( Fig. 2A ). Compression was highest for the 45flex posture with both night vision goggles and counterweight (315.8 ± 67.5 N), which was significantly higher than the night vision goggles (260.0 ± 33.4 N; Cohen’s d = 1.89), helmet (244.1 ± 40.4 N; Cohen’s d = 2.43), and no helmet conditions (213 ± 44.7 N; Cohen’s d = 3.47) for that posture (P < 0.05). A similar trend was observed in the extended posture, where the counterweight (242.1 ± 26.2 N) resulted in significantly more compression compared to the helmet (183.0 ± 24.1 N; Cohen’s d =1.99) and no helmet conditions (152.2 ± 24.1 N; Cohen’s d = 3.05). For extension, there was a trend for the counterweight to increase compression compared to the night vision goggles (205.4 ± 41.7 N), but this trend was not statistically significant (P = 0.082). In lateral bending, all helmet conditions resulted in increased compression from no helmet (202.5 ± 28.2 N) by an average of 58.8 N (Cohen’s d = 1.82, 1.98, 2.23 for hOnly, hNVG, and hNVG+CW, respectively). In a neutral posture, the compression from the night vision goggles (164.2 ± 3.7 N) was significantly greater than the no helmet condition (116.6 ± 5.7 N; Cohen’s d = 1.62).

View Figure

• Download as Powerpoint
• Download Figure

There was a significant interaction effect between posture and helmet condition on anteroposterior shear (Wald χ² = 26.14, df = 12, P = 0.01026). In all cases, the net force acting on the joint was directed posteriorly, indicated by the positive forces in Fig. 2B . In a neutral posture, anteroposterior shear is on average 4.9 ± 1.2 N, which was similar to a flexed posture (5.0 ± 1.6 N). In the lateral bending condition, both the helmet only and NVG+CW conditions resulted in significantly more posterior shear (32.9 ± 9.2 N and 34.0 ± 6.2 N, Cohen’s d = 0.40 and 0.43, respectively) compared to the no-helmet condition (21.2 ± 10.3 N).

Like the previous two outcome measures, there was a significant interaction effect between posture and helmet condition on mediolateral shear (Wald χ² = 38.3, df = 12, P = 0.00014; Fig. 2C ). In a neutral posture, for all helmet conditions, the mediolateral shear was almost identically zero (within 1 N). In the lateral bending condition, the counterweight significantly reduced the magnitude of mediolateral shear from 3.3 ± 1.6 N to 0.4 ± 1.0 N compared to the NVG condition (Cohen’s d = −2.79). Overall, the magnitudes of mediolateral shear tended to be low, below 10 N.

The flexion moment also exhibited a significant interaction effect between helmet and posture configuration (Wald χ² = 87.3, df = 6, P < 2 × 10⁻¹⁶). The night vision goggles induced a 0.83 ± 0.15 Nm increase in the flexion moment over the helmet alone (0.89 ± 0.08 Nm; Cohen’s d = 1.87) in a neutral posture, which was entirely negated by the adoption of the counterweight ( Fig. 3 ). In a flexed posture, the NVGs and counterweight induced a larger flexion moment (4.4 ± 0.6 Nm and 4.3 ± 0.5 Nm, Cohen’s d = 1.85 and 1.62, respectively) over just the helmet alone (3.7 ± 0.4 Nm). The counterweight accounted for the largest change in extensor moment in extension, amounting to −2.2 ± 0.4 Nm compared to no helmet (−1.2 ± 0.6 Nm, Cohen’s d = 2.51).

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

This study examined changes in the magnitude of joint reaction forces in the middle cervical spine with respect to postural and helmet configurations. The aim was to quantify the loads experienced by helicopter pilots in deviated postures while wearing a helmet, and night vision goggles with and without a counterweight. Overall, we found a nonlinear relationship between head-supported mass and resulting joint compression that interacted with posture. Added head-supported mass led to a disproportionate increase in joint loads. For instance, donning a 12.7 N helmet resulted in an increase of 27.2 N of compression—more than twice the weight of the helmet. This effect compounds with posture, for example, the largest magnitude of compression was with the NVG and counterweight system in flexion for a grand total of 315.8 N.

This investigation highlights that posture has a substantial influence over the joint loads, more so than the helmet center of mass location. To understand why, consider the displacement of the helmet’s center of gravity with the addition of night vision goggles and with the counterweight (Table I). These amount to 2–3 cm of either anterior or posterior displacement, in a neutral posture, relative to the C5-C6 joint center. For comparison, the same anterior-posterior displacements can be achieved with only 15° of flexion or extension, respectively. Therefore, for the helmet designs explored in this investigation, there exists postures for which the helmet center of mass is displaced to a greater extent than the initial helmet design. More to this point, in a neutral posture the night vision goggle weight induced a 0.83 Nm neck flexor moment which vanished in 30° of extension (Fig. 3). Conversely, in line with its design philosophy, the helmet-counterweight system induced an extensor moment in a neutral posture. In 45° of flexion, however, the center of mass of the system moves anterior to the C5/6 joint center ( Fig. 4 ), inducing a 4.4 Nm flexor moment, the same moment as the NVG condition. This is due to a tradeoff between increased helmet weight and decreased moment arm in this posture. However, for extension, the addition of the counterweight increased both the moment arm to the center of gravity of the helmet and the weight of the helmet system, resulting in a 2.2 Nm extensor moment. These moment changes translated to increased compression at this joint level with the addition of night vision goggles or the counterweight in both flexion and extension. This result highlights a challenging issue in helmet design: the mobility of the human neck. Helmet designs based on repositioning the center of mass of the helmet system may reduce the gravitational moment in one posture but may exacerbate it in another.

View Figure

• Download as Powerpoint
• Download Figure

Of the postures evaluated, a neutral position elicited the smallest joint reaction loads, approximately 116.5 ± 5.6 N of compression and 4.1 ± 1.1 N of posterior shear. Addition of the helmet, of weight 12.7 N, correspondingly increased the compression to 143.7 ± 4.3 N; a difference more than twice the actual weight of the helmet. This amplified increase in compression is due to muscle activity required to balance the gravitational moment. To this point, the NVG system, an added weight of 7.8 N, increased the compression further to 164.2 ± 1.3 N, a difference almost three times the weight of the NVGs. Counterintuitively, addition of a 6.2-N counterweight decreased the compression to 154.5 ± 4.6 N, simply due to the repositioning of the center of mass. These trends are similar in the lateral bending conditions, where donning a helmet with or without NVGs significantly increased joint compression from 202.4 ± 9.4 N to 261.3 ± 12.2 N. Overall, deviations from a neutral posture correspondingly increased the joint load, both in terms of compression and shear. This is in agreement with previous studies of helmeted individuals, where a more flexed posture has been repeatedly associated with a larger restorative moment requirement. ^13,35,37 Interestingly, cervical spine flexion, a known risk factor for the development of chronic neck pain, ^3,4 resulted in the largest magnitudes of compression: 315.8 N with a counterweight and night vision goggles—the highest compression examined in this study. This compression was 102.3 N higher than the no-helmet condition, nearly four times the magnitude of the total added weight of 26.7 N. Coincidentally, pilots frequently adopt flexed postures when they are scanning the ground for possible obstructions or the instrumentation clusters. ^24,38

The results of this study agree well with other literature on the topic. Since the counterweight did not reduce the mechanical parameters evaluated here, it is not surprising that its adoption into the military did not result in a corresponding decrease in reports of neck pain between 2004 and 2014. ^7,43 Forde et al. ¹³ used a simple inverse dynamics model to estimate the cumulative change in joint reaction forces and moments in helicopter pilots over flights. They found that pilots wearing NVG spent significantly more time in a flexed posture than those without, which translated into cumulative load values that were significantly higher on night flights than day flights. Similarly, in this study, a flexed posture was generally associated with greater loads that were magnified with the inclusion of NVG and counterweight. Over an entire night flight this would have manifested itself as an increase in all measures of cumulative load exposure. The compressive forces reported here with the NVG deployed are three times that reported in Forde et al., ¹³ while the anteroposterior and mediolateral shear values are slightly lower; this is attributable to the inclusion of muscular forces in this model.

Dibblee et al. ¹¹ designed and tested a novel spring at the back of the helmet to produce a supplementary extension moment without increasing the mass of the head-system. They found a significant decrease in measures of fatigue, decreased median power frequency, and ratings of perceived exertion when their intervention was used, and an increase in these measures when the counterweight was used over just the helmet alone. However, these investigators did not measure joint mechanics in their evaluation. Intuitively, the loading scenario of a posteriorly mounted pretensioned spring should be similar to the counterweight as both intend to provide a downward force at the head to counter the weight of the NVG. Future investigations would be needed to determine how this device changes joint mechanics. The compression and shear values reported here in a nonhelmeted condition agree very closely with an early model by Snijders et al. ³⁵ This was a relatively basic model of the cervical spine, partitioning it into three segments and including only 35 muscle elements, compared to the 218 used in this study. They later applied their model to the ergonomics of F-16 pilots in the Dutch Air Force and came to a similar conclusion: the helmet mass contributes significantly to the compression and shear loads in the cervical spine. ²¹

It should be noted that the magnitudes of compression and shear observed here are only 10% of what is required to acutely fail a functional spinal unit: it takes 3.9 kN of compression or 500 N of shear for failure to occur. ^32,34 Because these loads are so low in an acute scenario, it is rational to suspect that one source of the pain is muscular. Indeed, reducing muscular demands seems to be the current direction the Canadian Armed Forces is taking. ⁴³ However, muscular pain would not explain why reports of neck pain persisted after the introduction of the counterweight, as the counterweight has been shown to reduce the metabolic demand of the neck extensors. ^1,20 If the joint loads are responsible for the chronic neck pain experienced by most helicopter pilots, then it obviously implicates a mechanism of cumulative load rather than an acute loading event. This chronic pathway also explains why RCAF pilots with more flight time report more neck pain. ¹ Gooyers and Callaghan ¹⁶ used an in-vitro porcine cervical spine model to show that loads as low as 10% of the ultimate compressive tolerance are enough to significantly alter the surface geometry of the intervertebral disc in response to a 0.5-Hz loading regime in a flexed posture over a 2-h loading protocol. Coincidentally, compression in helmeted conditions and flexed postures are close to the 10% of failure strength (Fig. 2), potentially implicating the intervertebral disc as a source of pain in helicopter pilots. Future studies are necessary to fully determine if these low-magnitude loads sustained over an extended duration are sufficient for the development of osteoligamentous injury, or if the reduction of load in the helmet is an effective means of mitigating the pain experienced by helicopter pilots.

Chronic neck pain is a multifaceted issue attributable to psychosocial, physiological, and mechanical factors. ^10,20,31 Nevertheless, this study documented that the presence of a helmet and NVG significantly alter the mechanical exposure of helicopter pilots, which the counterweight fails to address comprehensively across the range of postures adopted during flight.

The most significant limitation in this investigation was the omission of whole-body vibration (WBV), which helicopter aircrew endure. Exposure to WBV, especially near the resonant frequency of the spine, is very strongly associated with the development of low back and neck pain ^30,40,41 and microscopic tissue damage. ¹⁸ Further, the posture of operators, an attribute explored extensively in this investigation, is known to influence the mechanical response of individual functional spinal units to vibration. ^5,17 Additionally, work with in vivo participants has shown that individuals exposed to WBV co-contract their spinal musculature, which stiffens and stabilizes their spines; ¹⁹ naturally, the elevated muscle activity would likely increase the joint loads. Theoretically, since the vibration modes of a mechanical system are related to both its mass and stiffness matrices, ¹⁵ we hypothesize that the magnitude of head-supported mass from a helmet would interact with the stiffness from co-contraction to alter the cervical spine’s vibration behavior. Unfortunately, this analysis fell outside of the scope of the current investigation.

The study was carried out on a healthy male population, which may not be representative of the helicopter pilots who work for the Canadian military. Additionally, subjects carried out static postural holds which are also likely not representative of the ballistic movements helicopter pilots need to undertake as they control their aircraft, nor all the static postures assumed by operators. Combined flexion with axial rotation, for example, is also a very common posture which was not explored in this investigation. ^13,36 These intermittent bursts during dynamic tasks would plausibly produce compression and shear forces greater than those reported here, and it is unclear how the presence of the counterweight would either mitigate or exacerbate those forces. Similarly, subjects replicated these postures in a controlled laboratory environment, which is different from the noninertial reference frame that is the cockpit of an active CH-146. Finally, no attempt was made to quantify the resulting inertial forces that would be acting on the head: the Coriolis, centrifugal, or Euler forces, which are proportional to the mass of the head and helmet system.

The model suffers from limitations as well. In terms of kinematics, the model linearly partitions the overall head-neck angle among the intervertebral joints. This is a necessary model assumption but simplifies the number of different configurations the cervical spine can adopt while maintaining the same external head-neck angle. ^28,39 In terms of kinetics, there are a number of limitations germane to the Hill-type muscle model, most notably its failure to characterize force production capacity of submaximal eccentric contractions. ^22,44

In conclusion, this study quantified the loads on the middle cervical spine in a variety of posture and helmet conditions. We hypothesized that joint reaction loads would increase proportionately with the magnitude of head-supported mass, and the proportionality would be posture-specific. Supporting our hypothesis, compression at C5-C6 increased proportionately with the weight of the helmet and night vision goggles. For instance, donning a 12.7 N helmet correspondingly increased the compression by 27.2 N. Further, addition of the NVG system, an added weight of 7.8 N, increased the compression another 20.5 N. This amplification can be exacerbated by posture; for instance, a flexed posture with helmet, NVG, and counterweight systems with a total weight of 26.7 N, exhibited a 102.3 N increase in compression compared to the same posture without a helmet. These findings supported our hypothesis that repositioning the center of gravity leads to posture-specific responses, with beneficial responses in neutral head-neck positions, but detrimental loading in nonneutral postures. This result may be of interest to designers of helmets, night vision goggles, and counterweights. The increased moment demands generally translated into greater compression and anteroposterior shear. For every static posture and helmet conditions, the compression and anteroposterior shear magnitudes were generally lower than 10% of the ultimate tolerance, which we hypothesize to implicate a cumulative loading pathway for the development of chronic neck pain.