Low-back injury and pain compromises performance in many athletes. Unfortunately, what is known about basic back function lags implementation for both prevention of troubles and their rehabilitation. For example, periods of sitting are linked with low-back troubles (3,7). Also, unsupported sitting is characterized by significant lumbar flexion (2). Prolonged flexion is known to creep the viscoelastic ligaments and disk (9), temporarily compromising spine stability and causing redistribution of the disk nucleus (6). In addition, it is well known that disk height determines ligament slackness, which in turn influences range of motion. Further, there may be an interaction between the increasing length and slackness of the ligaments surrounding the disk and torsional viscosity of the nucleus during static loading that could modulate stiffness during periods of no motion. Consideration of this knowledge obtained from basic science implies that athletes who warm-up to play and then “sit on the bench” for rest may be compromising their low-back health and increasing injury risk. This study was designed to assess whether this practice changes the mechanics of the low back.
In most team sports, players perform warm-up and stretching procedures in preparation for a game. After this warm-up, the players who do not start in the game generally proceed to “sit on the bench” and are often substituted into the game after an indefinite amount of time. A great majority of the time, the bench that is used by players provides no back support and sits very low to the ground. As well, players in sports such as basketball and volleyball may sit on the bench for half an hour or more at a time.
It was hypothesized that a warm-up and stretching procedures will decrease lumbar spine stiffness and increase lumbar spine range of motion for the athletes tested. It was also hypothesized that a period of inactive bench sitting after such a warm-up will negate any measurable benefits of the warm-up routine. Specifically, one would expect the spine to increase in overall lumbar stiffness and decrease in lumbar range of motion. A group of varsity volleyball players were studied given their general practice of warming up and sitting on the bench for substantial lengths of time before the immediate resumption of play.
Nine healthy male subjects were recruited from the university varsity volleyball population (age = 23.2 ± 3.6 yr, weight = 85.9 ± 9.5 kg, height = 1.86 ± 0.08 m. Eligibility requirements included no reports of chronic or disabling low-back pain within 1 yr before testing. Each subject signed information and consent forms approved by the Office of Research Ethics at the University of Waterloo.
Before beginning any of the stiffness tests each participant performed maximum voluntary range of motion tasks for flexion, extension, lateral bend and axial twist. These motions were not forced, and a recovery time was provided to minimize any possible impact on subsequent tests of stiffness. In addition, measurements of participant height and weight were taken. Each of three stiffness tests was performed for four movements (spine flexion, extension, right lateral bend, and clockwise rotation, tested in random order) with three trials collected in sequence per movement for a total of 36 trials per subject. The first stiffness test was performed initially upon entering the lab on each subject. Each participant then proceeded to warm-up for 30 min adhering to the “typical” warm-up procedure, the specifics of which are presented in Figure 1. After this, they were tested for stiffness again, before being instructed to sit on a standard team bench for 30 min. While sitting, 10 s of lumbar kinematic data was collected every 5 min. After this period of bench rest, each subject was tested for stiffness a final time.
Low-back flexion-extension and lateral bending stiffness were measured on a jig that fixed the lower body at the pelvis on an immovable platform and fixed the upper body to a support that rolled on a virtually frictionless “floating” bearing system that allowed unrestricted planar “X-Y” motion (Fig. 2). Axial twist stiffness was measured with participants standing on a turntable with their torso fixed at the ribcage (Fig. 2). The equipment used in securing the participant for all movements was used in a previous study investigating passive stiffness measurement of the lumbar spine by McGill et al. (10).
Angular motion of the lumbar spine was isolated and measured using a 3-SPACE Isotrak system (Polhemus Navigation Systems, Colchester, VT). All signals were A/D converted at 60 Hz. This system consisted of a source and a sensor module where the source was secured over the sacrum of the subject by straps around the hips and between the legs. The sensor module was placed on the subject’s back at the level of T12 and secured by a strap around the torso (Fig. 3). The system measured three-dimensional motion with an accuracy of ± 0.3° (10).
In all movements, a force was applied to the top of the platform supporting the upper body (or to the edge of the turntable in axial twist) via a steel cable, along a line perpendicular to the long axis of the body. Tension in the cable was measured with a load cell and bending torque was calculated using the perpendicular distance from the cable attachment to approximately the level of L4/L5. At all times during the stiffness testing, the welfare of the subject was the first priority, such that torque was applied until the subject verbally indicated that they did not wish to proceed further. Each trial lasted 20 s with the maximum torque being reached around the 15- to 18-s mark.
Bipolar pairs of myoelectric electrodes (Ag-AgCl) were adhered to the skin over the left and right external oblique (15 cm lateral to the umbilicus). The signal from these electrodes was played over a stereo speaker system in order to provide the subjects with audio feedback about the activation level of their trunk musculature during the tests. Certain random bursts of muscle activity were unavoidable, especially upon nearing the end range of motion (possibly reflexive), but trials with extended periods of muscle activity were recollected.
Raw data from the load cell (A/D units) were converted to Newtons by multiplying by a conversion factor for the load cell. These values were then converted to torque (N·m) by multiplying each value by the moment arm that was measured for each trial.
The concept of stiffness was examined using two different approaches. The first was to measure the amount of torque needed to pull the subject to a given angle in each of the three conditions (pretest, post exercise, and post rest). The second method was to look at the slope of the torque (N·m) versus angle (deg) curve at a given angle. The slope of this curve gives a direct measure of stiffness (in N·m/deg) at a particular point. An increase in torque or slope at a given angle from one condition to another is indicative of an increase in stiffness of the lumbar spine at the angle being compared. Similarly, a decrease in torque or slope shows a decrease in stiffness at that angle. Both torque and slope were examined at 80% and 90% of the maximum angle taken from the post rest condition of each subject. These angles were chosen because injury potential is greater near the end range of motion where the ligaments and intervertebral disk are stressed. Additionally, comparisons of both slope and angle were made at 80% and 90% of the maximum torque taken from the post rest condition of each subject. Torque and angle analyses served as more of a global measure of stiffness change. For example, a higher torque at a given angle in one condition was indicative of a change in stiffness from one condition to the other but not necessarily at the point being compared.
A comparison was made between pretest and post exercise curves to determine whether the warm-up had an effect on lumbar stiffness using simple “walking to the lab” as a base comparator. All measurements were taken at least 3 h after rising from bed to avoid the stiffness associated with imbibed disks (4). The comparison between post exercise and post rest curves was made to determine whether bench rest had an effect on lumbar stiffness in an individual that had already warmed up. Finally, the comparison was made between pretest and post rest curves to see whether the interaction of a warm-up followed by bench rest would cause an increase in stiffness of the lumbar spine.
The post rest condition was chosen as the base for comparison because it formed the hypothesis that bench sitting would lead to an increase in stiffness from the postexercise condition and therefore the maximum torque and angle of the post rest curve should be less then those of the postexercise curve. In this way, data points from the other test conditions were assured. For example, if the torque and angle values were chosen from the postexercise curve, there may not have been any corresponding values at those levels in either the pretest or the post rest curves because they were smaller values.
A single point was used for the comparisons because the linearization of data would have resulted in the loss of important information about the end range. Changes in the mid-region of the neutral zone were not as important as changes in the end range for athletic application, as the end range has important implications for injury mechanics and performance. A one-way ANOVA was used in the comparison of values from pretest to post exercise to post rest.
A warm-up followed by bench rest was found to increase stiffness of the lumbar spine in some movements but not in others. Specifically, in extension, a greater amount of torque was required to pull subjects to a given angle in the post rest condition compared to the post exercise condition (Table 1). Also, the slope of the torque by angle curve at a given angle was found to be steeper in the post rest condition than in the post exercise condition in extension. Similarly, in lateral bend, the slope of the torque by angle curve at a given torque was found to be steeper in post rest than in post exercise (Table 2).
A significant difference was seen between pretest and post rest torque at a given angle in lateral bend, and also between pretest and post rest slope (of the torque by angle curve) at a given angle in both extension and lateral bend (an example of these differences in torque and slope at a given angle between post exercise and post rest as well as between pretest and post rest is provided in Fig. 4). There were no significant pretest to post exercise differences observed for any movement. Additionally, both flexion and axial twist measurements did not produce any significant differences between tests (all significant means and standard deviations are organized in Tables 1 and 2).
During the “rest” task all participants assumed a flexed posture while sitting with a lumbar spine angle in the sagittal plane having an average ± SD of 50.32° ± 7.65°. Clearly, all subjects sat with a large amount of lumbar flexion (portrayed in Fig. 5).
Subjects demonstrated some variability in response to the various conditions. For example, some appear to increase stiffness with a warm-up followed by rest whereas others do not. An example is shown in Figure 6.
With respect to the hypotheses, an increase in stiffness was observed from pretest to post rest and from post exercise to post rest in extension (Table 1) and lateral bend (Table 2). There was no observed decrease in stiffness as a result of warm-up and no observed change in range of motion between any of the conditions. Therefore, it would appear that a warm-up on its own does not substantially alter spine stiffness but a warm-up followed by bench rest results in an increase in stiffness of the lumbar spine in some motions. Those with symptomatic backs in particular may benefit from addressing the additional stiffness from bench sitting before the resumption of play.
The interpretation of these data is limited to varsity-level volleyball players who are physically fit and do not conform to the demographics of the usual population (e.g., height = 1.86 ± 0.08 m) but may also better represent athletes such as basketball players. An attempt was made to provide EMG of the participant’s trunk muscles as biofeedback to ensure a passive response, although small amounts of muscle activation occurred occasionally. These trials were discarded and subjects were able to learn to relax. In addition, the randomized presentation of conditions controlled for any ordering effect but may have influenced the variability in subject response. Finally, there may have been small amounts of motion between the apparatus and subjects that were minimized by tightly securing the subjects with straps.
Of interest is the fact that a 26-min warm-up with specific stretches for the low back and legs did not lead to a significant decrease in stiffness. Perhaps the stretching at the end of the warm-up contributed to this finding. Even though the warm-up would have led to an increase in muscle temperature and a reduction in muscle viscosity (11), it has been shown that activity itself does not always result in a decrease in stiffness (8). Additionally, the stretching exercises provided participants with a chance to “cool down” as their body temperature returned to normal. This could have counteracted any significant effect that the warm-up may have had on decreasing stiffness. There was also a great deal of individual variability seen in the participants of this study (Fig. 6). Some subjects responded to the warm-up with increasing stiffness whereas others with more compliance. Similarly, some participants were found to be less stiff after bench rest whereas others were found to be stiffer. It is also interesting to note that an increase in stiffness across subjects was found only in extension and lateral bend, and not in flexion or axial twist. This phenomenon may be a function of the testing protocol. In the twist condition, participants were secured in a different jig than any other movement and the torque was applied to their feet, not their torso like the other three conditions (Fig. 2). With regard to the flexion stiffness test, the apparatus did not have a large enough “floating” surface to pull some participants through their entire range of motion, limiting the amount of data that could be collected in this condition.
The stiffness data calculated in this study suggest that there is an increase in lumbar spine stiffness as a result of prolonged bench rest after a warm-up. As well, there is some degree of individual variability seen in the response to such activity. Because a period of extended sitting is common to many team sports, athletes and coaches alike should be aware of this stiffness change as well as the potential threat to low-back health and the possibility of decreased performance that may accompany this behavior. In addition, sitting with a flexed spine is not only linked with disk herniation (5) but is also an injury mechanism (1,12), suggesting this posture is contraindicated for athletes suspected of being at risk for this condition. These results may be generalizable to other team sports such as basketball, field hockey, soccer, and ice hockey in which players sit on a team bench before playing. Further studies in this area may include research into how much time an athlete needs to sit on a bench to cause an increase in lumbar spine stiffness, together with research into alternative postures or tasks for “bench players” to assume while waiting to play.
The authors gratefully acknowledge the financial support of the Natural Sciences and Engineering Research Council of Canada (NSERC).
Address for correspondence can be sent to: Professor S. McGill, Occupational Biomechanics Laboratories, Department of Kinesiology, University of Waterloo Ontario, Canada, N2 L 3G1; E-mail: email@example.com.
1. Callaghan, J. P., and S. M. Mcgill. Intervertebral disc herniation: studies on a porcine model exposed to highly repetitive flexion/extension motion with compressive force. Clin. Biomech. 16: 28–37, 2001.
2. Callaghan, J. P., and S. M. Mcgill. Low back joint loading and kinematics during standing and unsupported sitting. Ergonomics 44: 280–294, 2001.
3. Frymoyer, J. W., M. H. Pope, M. C. Costanza, J. C. Rosen, J. E. Goggin, and D. G. Wilder. Epidemiologic studies of low-back pain. Spine 5: 419–423, 1980.
4. Gunning, J. L., J. P. Callaghan, and S. M. Mcgill. Spinal posture and prior loading history modulate compressive strength and type of failure in the spine: a biomechanical study using a porcine cervical spine model. Clin. Biomech. 16: 471–480, 2001.
5. Kelsey, J. L. An epidemiological study of the relationship between occupations and acute herniated lumbar intervertebral discs. Int. J. Epidemiol. 4: 197–205, 1975.
6. Krag, M. H., R. E. Seroussi, D. G. Wilder, and M. H. Pope. Internal displacement distribution from in vitro loading of human thoracic and lumbar spinal motion segments: experimental results and theoretical predictions. Spine 12: 1001–1007, 1987.
7. Lengsfeld, M., A. Frank, D. L. van Deursen, and P. Griss. Lumbar spine curvature during office chair sitting. Med. Eng. Phys. 22: 665–669, 2000.
8. Magnusson, S. P., P. Aagaard, B. Larsson, and M. Kjaer. Passive energy absorption by human muscle-tendon unit is unaffected by increase in intramuscular temperature. J. Appl. Physiol. 88: 1215–1220, 2000.
9. Mcgill, S. M., and S. Brown. Creep response of the lumbar spine to prolonged full flexion. Clin. Biomech. 7: 43–46, 1992.
10. Mcgill, S. M., J. Seguin, and G. Bennett. Passive stiffness of the lumbar torso in flexion, extension, lateral bending, and axial rotation. Spine 19: 696–704, 1994.
11. Shellock, F. G., and W. E. Prentice. Warming-up and stretching for improved physical performance and prevention of sports-related injuries. Sports Med. 2: 267–278, 1985.
12. Wilder, D. G., M. H. Pope, and J. W. Frymoyer. The biomechanics of lumbar disc herniation and the effect of overload and instability. J. Spinal Disord. 1: 16–32, 1988.
Keywords:©2002The American College of Sports Medicine
LOW-BACK HEALTH; LUMBAR SPINE STIFFNESS