Share this article on:

Central Activation and Force-Frequency Responses of the Lumbar Extensor Muscles


Medicine & Science in Sports & Exercise: July 2009 - Volume 41 - Issue 7 - p 1504-1509
doi: 10.1249/MSS.0b013e31819b3607
Special Communications

Purpose: To determine the central activation and force-frequency properties of a muscle group in which these data have not previously been reported (i.e., the lumbar extensors).

Methods: Fifteen young healthy participants were tested. Maximal voluntary isometric contraction (MVIC) of the trunk was collected using a custom apparatus with a cable-mounted load cell. Central activation was determined by delivering a supramaximal 100-ms, 100-Hz train to the lumbar muscles during the MVIC. The MVIC and the stimulated forces were used to calculate the central activation ratio (CAR) for each subject using the formula CAR = MVIC force / (MVIC + stimulated force), with a value of 1 indicating full central activation. Force-frequency relationship of the muscle group was determined by stimulating the muscles every 10 s with two 500-ms trains of the following frequencies: 1, 5, 10, 20, 40, 60, 80, and 100 Hz. The force-frequency data were fit with a four-parameter Hill equation to obtain the frequency at which 50% of the range of forces generated by the various stimulation trains is produced (F 50) and the Hill coefficient, which indicates the steepness of the linear portion of the relationship.

Results: Mean MVIC was 345.4 N (SD = 126.7), and mean CAR was 0.95 (SD = 0.06). The force-frequency data showed a mean F 50 of 16.40 Hz (SD = 3.15) and a mean Hill coefficient of 2.21 (SD = 0.50).

Conclusions: Central activation and force-frequency testing of the lumbar extensor muscles is feasible, and the data reported here represent, to our knowledge, the first of their kind in this muscle group.

School of Physical Therapy, Ohio University, Athens, OH

Address for correspondence: James S. Thomas, P.T., Ph.D., School of Physical Therapy, Ohio University, W277 Grover Center, Athens, OH 45701; E-mail:

Submitted for publication October 2008.

Accepted for publication January 2009.

There are numerous studies that use neuromuscular electrical stimulation to assess the degree of central activation (4,12,24,26,28) and the force-frequency relationship (FFR) (1,5,14), two elements that relate to the production of muscular force and muscle fatigue. However, these studies have been primarily limited to the muscles of the upper and lower extremities and have not assessed these elements for the lumbar extensor muscles. Lumbar extensor muscle force and fatigability are thought to be important factors in prevention of, and recovery from, low-back pain (LBP) (7,16,22,32). Because changes in neuromuscular activation of the limb muscles have been identified in both acute and chronic clinical populations, it is reasonable to think that similar changes could occur in the trunk muscles. Thus, development of a method to examine central activation and FFR could assist in the determination of mechanisms contributing to deficits in individuals with LBP.

Central activation is frequently defined as the ability of the CNS to activate a given muscle. Assessment of the degree of central activation is typically performed using a combination of volitional contraction and electrical stimulation. One commonly used method for assessing central activation involves delivering a short, high-frequency train of supramaximal pulses to the muscle(s) of interest via the peripheral nerve during an isometric maximum voluntary contraction (MVC) (12). The key assumption of this method is that the electrical stimulation will increase force production in any muscle that is not being fully recruited or activated with a sufficiently high degree of rate coding.

The determination of central activation in various limb muscles has been proven useful for revealing activation impairments in several conditions (e.g., fatigue) and clinical populations. Central activation failure has been shown to significantly contribute to muscle weakness in clinical populations, including persons with osteoarthritis, cerebral palsy and those status post knee arthroplasty (9,19,27), and to negatively influence the relationship between muscle strength and physical function (9). Because these activation deficits contribute to impairment of muscle force production, it has been suggested that rehabilitation interventions that facilitate activation (e.g., biofeedback and neuromuscular electrical stimulation) may prove more useful in these populations than traditional exercise protocols (19).

The FFR describes how the force of activated motor units is modulated by varying the discharge rate. The FFR varies with many factors, including muscle group, habitual use, and age, but the general sigmoid nature of the relationship is preserved across in vivo whole muscle (5), single motor unit (10), and single fiber models (20). Although not used in clinical studies to the same extent as the central activation ratio (CAR), acute changes in the FFR occur with muscle fatigue. Typically, a rightward shift of the FFR occurs, such that higher activation frequencies are needed to maintain thesame relative submaximal level of force. This shift is indicative of the phenomenon of low-frequency fatigue and is thought to indicate an acute impairment of excitation-contraction coupling (34).

Lumbar extensor muscle force and fatigability are thought to be important factors in prevention of, and recovery from, low-back pain (LBP) (7,16,22,32). LBP has a lifetime prevalence rate of 60-90% in industrialized nations, and this prevalence rate is not diminished among those participating in sports (11). In addition, there is considerable evidence that athletes who participate in sports such as rowing, golf, and gymnastics have significant problems with LBP (6,8,23). Because changes in neuromuscular activation of the limb muscles have been identified in both acute and chronic clinical populations, it is likely that similar changes could occur in the trunk muscles. Thus, examination of the central activation and the FFR in the trunk muscles could assist in the determination of mechanisms contributing to deficits in individuals with LBP. Because such measurements have not been reported for the lumbar extensors, in either healthy or patient populations, it is important to first develop a method for testing neuromuscular activation of the trunk muscles in a healthy population. The present study describes a methodology for testing the central activation and the FFR in the lumbar extensors, demonstrates the feasibility of collecting these data, and provides preliminary data regarding baseline normal values that can serve to facilitate further research in clinical populations.

Back to Top | Article Outline


Fifteen participants (nine women, six men) with an average age of 24.6 yr (SD = 4.3) were tested in a custom apparatus designed to measure lumbar extensor force with a cable-mounted load cell sampling force at 1 kHz. The participant's average height was 1.7 m (SD = 1.3), and their average weight was 72.2 kg (SD = 14.6). Thirteen participants completed the entire experimental procedure, and those data are presented here. All subjects gave written, informed consent before the administration of any experimental procedures. The study was approved by the institutional review board of Ohio University.

Back to Top | Article Outline

Experimental apparatus.

A custom apparatus was constructed from metal framing (Unistrut, Wayne, MI) for the purpose of stabilizing the subjects in modified kneeling position (Fig. 1). Great care was taken to immobilize the pelvis using a custom clamp as shown in Figure 1. A customized harness was placed over the trunk and held in position with inextensible nylon straps. A load cell (Load Cell Central, Monreton, PA) was attached to the harness via a steel cable, which was run through a two-pulley system affixed to the testing apparatus. The output from the load cell was amplified with an OM-19 (Load Cell Central) sampled at 1000 Hz using customized LabView software and stored for subsequent analyses. The force data were filtered with a fourth-order zero-lag Butterworth filter with a 6-Hz cutoff frequency. Before recording, the slack was taken out of the load cell-pulley system.



Back to Top | Article Outline

Muscle testing protocol.

Participants were tested in a modified kneeling position with the lumbar spine in neutral and the pelvis immobilized. A pair of 2-inch electrodes (Axelgaard, Fallbrook, CA) was placed over each (right and left) lumbar extensor muscle mass and connected to a DS7A constant-current stimulator (Digitimer, Hertfordshire, UK). The stimulators were synchronously triggered by customized LabView software. Before performing the maximal voluntary trunk extension, participants were given several practice trials where they were verbally encouraged to pull back on the harness as forcefully as possible with the trunk while one experimenter monitored the participant to ensure that they did not raise their buttocks off the seat. If raising occurred, we assumed that gluteal substitution was taking place and we repeated the trial. Participants then performed three maximal voluntary isometric contractions (MVIC) with verbal encouragement. For the FFR testing, stimulus amplitude was set so that a 100-ms, 100-Hz train produced 50% of MVIC. The FFR was determined by stimulating the muscles every 10 s with 500-ms trains of the following frequencies: 1, 5, 10, 20, 40, 60, 80, and 100 Hz, with the stimulation sequence delivered first in ascending order and then descending order. After determining the FFR, maximum stimulated force was determined by increasing the current through each stimulator until no further increase in force was elicited by a single pulse. For supramaximal stimulation, the stimulus amplitude was increased by 15% of this value. We then had the participant perform three MVIC. Force output was monitored in real time, and when the participant reached their MVIC (determined from the earlier trials), the experimenter manually triggered the superimposed burst of supramaximal stimulation (100 ms, 100 Hz). Participants were given 5 min of rest between each trial.

Back to Top | Article Outline

Data analysis.

The MVIC was taken as the highest force produced during each of the three trials. For each trial, the MVIC and the stimulated forces were used to calculate the central activation ratio (CAR) for each subject using the formula CAR = MVIC force / (MVIC + stimulated force), with a value of 1 indicating full central activation (Fig. 2). The CAR of each participant was defined as the average CAR from the three trials. The mean peak force responses to the two trains at each stimulation frequency were used to generate the FFR for each subject. Each FFR was fit with a four-parameter Hill Equation (SigmaPlot 9.0, Systat Software) to obtain the F 50. The F 50 is defined as the frequency at which 50% of the range of forces generated by the various stimulation trains (e.g., 1,5,20, 40, 60, 80, and 100 Hz) is produced. The fitting procedure also provides the Hill coefficient, which indicates the steepness of the linear portion of the relationship. Although several investigators have determined both the F 50 and the slope of the FFR using linear interpolation (1,5,15,30), fitting the FFR data with the Hill Equation provides more objective values for these parameters and has been performed in other muscle groups (18,25). All data are reported as means (SD).



Back to Top | Article Outline


The mean MVIC force for the subjects was 345.4 N (SD= 126.7). The force responses during the central activation testing were similar to those seen in other muscles (Fig. 2), and the CAR values were generally high, with a mean of 0.95 (SD = 0.06). The FFR exhibited the typical sigmoid shape (Fig. 3), which exhibited a mean F 50 of 16.40 Hz (SD = 3.15) and a mean Hill coefficient of 2.21 (SD = 0.50).



Of the 15 subjects tested, two experienced an episode of vasovagal syncope and were unable to complete the experimental protocol. Neither of the episodes occurred during stimulation. Rather, the two subjects reported the onset of symptoms (light headedness, flushing) during the rest period between stimulation trains. Interestingly, neither subject complained of pain during the delivery of the stimulation trains. Finally, both participants recovered within minutes, sustained no injuries, and reported no persistent adverse effects.

Back to Top | Article Outline


The present study describes the adaptation of established methods for assessing central activation and FFR for use in the lumbar extensor muscles. This process required several iterations, but the results demonstrate that collection of these data in the lumbar extensors is feasible. Furthermore, the data here represent, to our knowledge, the first of their kind in this muscle group.

Several prototypes of the experimental apparatus and the stimulation configuration were tested before settling on the final model with the pulley-mounted load cell shown in Figure 1. A key element in measuring CAR and FFR in this muscle group was obtaining adequate stabilization of the pelvis. Another necessity was using two separate stimulators to activate the left and the right lumbar extensors. Our initial attempts using a single unit were hampered by an inability to fully activate the target muscles, most likely because splitting the leads to allow activation of the two lumbar extensor muscle masses produced a current divider. Using two stimulators allowed us to deliver sufficient current and to adjust the stimulation to each side independently. Similar results could likely be obtained using a single stimulator with two independent channels.

Although we were initially surprised by two episodes of syncope during data collection, participants have been known to pass out during various electrical stimulation protocols (Stuart Binder-Macleod, personal communication). Close examination of the data from these two participants revealed that they had an atypical muscle response to increasing amplitudes of electrical current. In most subjects, low current amplitudes (i.e., 10-20 mA) produced low extensor torques, and increasing current amplitude produced rapid increases in torque. However, in the two participants who had the episode of syncope, their extensor torque did not increase as stimulation intensity increased. Furthermore, the absolute current delivered was no higher than that administered to the other participants. Previous animal studies clearly indicate the role of muscle contraction in moderating the effects of autonomic system discharge, and this could explain the syncope in these participants (13). We speculate that because these participants had only weak trunk muscle contractions, the concomitant stimulation of the autonomic nervous system was then amplified. Future studies in our laboratory will screen participants for risk of syncope by testing their responsiveness trunk muscles to electrical stimulation.

To assess central activation, we chose to determine the CAR, as described by Kent-Braun and LeBlanc (12). The CAR values reported here are similar to those reported in healthy participants in several other muscle groups (3,4,19,24,26,28,30). Another commonly used method for determining the degree of central activation is the twitch interpolation method, in which a supramaximal twitch is delivered to the muscle(s), with the muscle at rest and during an isometric MVC (17). The amplitude of the superimposed twitch is compared with the resting twitch. Although debate persists over the relative utility of the two methods (3,12), we chose to use the CAR technique because it has been shown to produce greater increases in force when superimposed on an MVIC than a twitch or doublet. Because we did not know what sort of response to expect with this novel application, we wanted to be sure we saw a difference in force with the stimulation. It has been established in the quadriceps muscles that the relationship between CAR and voluntary effort is not linear but rather is best fit by a second-order polynomial (28,29,31), and the exact model parameters likely vary with different population characteristics, such as age (29,31). The authors of these studies have recommended correcting the CAR by using the quadratic equation that best fits the voluntary effort-CAR relationship for a given population. By establishing the methods presented here, it will now be possible to determine whether such correction is necessary in the lumbar extensors and to compare the utility of the CAR to that of the twitch interpolation technique.

Impairments of the CAR have been shown to contribute to muscle weakness, independent of any effects of pain, in several clinical populations including patients who underwent knee arthroplasty (19) and individuals with osteoarthritis (9), cerebral palsy (27), and human immunodeficiency virus (26). The present results now offer the potential to test for such impairments in clinical LBP populations. Although testing of central activation might not be especially practical in an acute LBP population, it may provide great insight in studying recurrent and chronic LBP. It may be that the reason recurrence of LBP is as high as 50% (2) is that certain individuals are more susceptible to changes in neural activation of the trunk extensors after an initial episode, even after they have reported recovery from symptoms. For example, in a previous study of 88 participants who had fully recovered from an episode of LBP and were asymptomatic, there were still deficits in peak lumbar velocity and acceleration during reaching tasks (33).

The FFR of the lumbar extensors exhibited the general sigmoid form seen in other muscles. The F 50 values reported here are between those reported for the knee extensors (21.6 Hz) (25) and the dorsiflexors (14 Hz) (18) in studies that used the Hill equation to calculate the F 50. Other investigators who define the F 50 as simply the frequency at which 50% of the maximum force is produced report lower F 50 values in both the knee extensors (13-18 Hz) (1,5,30) and the ankle dorsiflexors (12 Hz) (21). If we use this latter definition for our data, we obtain a value of ∼10 Hz, lower than what has been reported for either muscle group, but certainly closer to the dorsiflexors. The difference in F 50 values is essentially the result of correcting for the offset in the force data (i.e., there is no stimulation frequency that produces 0 force).

Few data are available on the Hill coefficients of different muscle groups, but one laboratory has reported a value of ∼3.5 for the ankle dorsiflexors (18). Analysis of a data set of control participants used in a published manuscript (25) gave a value of ∼2.1 for the knee extensors (S. Binder-Macleod, personal communication), much closer to the value of 2.2 found here for the lumbar extensors. Interestingly, it has been suggested that muscle groups that depend more on rate coding to modulate force production exhibit steeper FFR (greater Hill coefficients) than those that use it less (15).

Back to Top | Article Outline


Central activation and force-frequency data represent important physiological parameters of skeletal muscle but have not been previously reported in this muscle group. Having demonstrated the feasibility of collecting these data in healthy participants, it will now be possible to directly assess the contribution of these factors in recurrent and chronic LBP and identify potential targets for rehabilitation strategies.

The authors thank Stephanie Gustwiller for her assistance in data collection. This work was supported by the National Institutes of Health, R01 HD045512 (Dr. Thomas). The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Allman BL, Rice CL. An age-related shift in the force-frequency relationship affects quadriceps fatigability in old adults. J Appl Physiol. 2004;96(3):1026-32.
2. Andersson GBJ. Epidemiological features of chronic low-back pain. Lancet. 1999;354(9178):581-5.
3. Behm DG, Whittle J, Button D, Power K. Intermuscle differences in activation. Muscle Nerve. 2002;25(2):236-43.
4. Belanger AY, McComas AJ. Extent of motor unit activation during effort. J Appl Physiol. 1981;51(5):1131-5.
5. Binder-Macleod SA, Halden EE, Jungles KA. Effects of stimulation intensity on the physiological responses of human motor units. Med Sci Sports Exerc. 1995;27(4):556-65.
6. Cole MH, Grimshaw PN. Trunk muscle onset and cessation in golfers with and without low back pain. J Biomech. 2008;41(13):2829-33.
7. Cooper RG, Stokes MJ, Sweet C, Taylor RJ, Jayson MI. Increased central drive during fatiguing contractions of the paraspinal muscles in patients with chronic low back pain. Spine. 1993;18(5):610-6.
8. Cupisti A, D'Alessandro C, Evangelisti I, Piazza M, Galetta F, Morelli E. Low back pain in competitive rhythmic gymnasts. JSports Med Phys Fitness. 2004;44(1):49-53.
9. Fitzgerald GK, Piva SR, Irrgang JJ, Bouzubar F, Starz TW. Quadriceps activation failure as a moderator of the relationship between quadriceps strength and physical function in individuals with knee osteoarthritis. Arthritis Rheum. 2004;51(1):40-8.
10. Fuglevand AJ, Macefield VG, Bigland-Ritchie B. Force-frequency and fatigue properties of motor units in muscles that control digits of the human hand. J Neurophysiol. 1999;81(4):1718-29.
11. Jacob T, Baras M, Zeev A, Epstein L. Physical activities and low back pain: a community-based study. Med Sci Sports Exerc. 2004;36(1):9-15.
12. Kent-Braun JA, Le Blanc R. Quantitation of central activation failure during maximal voluntary contractions in humans. Muscle Nerve. 1996;19(7):861-9.
13. Koba S, Xing J, Sinoway LI, Li J. Differential sympathetic outflow elicited by active muscle in rats. Am J Physiol Heart Circ Physiol. 2007;293(4):H2335-43.
14. Lee SC, Russ DW, Binder-Macleod SA. Force frequency relation of skeletal muscle. In: Binder MD, Hirokawa N, Windhorst U, Hirsch MC, editors. Encyclopedia of Neuroscience. New York: Springer; 2008 Vol. 2, p. 1608-11.
15. Macefield VG, Fuglevand AJ, Bigland-Ritchie B. Contractile properties of single motor units in human toe extensors assessed by intraneural motor axon stimulation. J Neurophysiol. 1996;75(6):2509-19.
16. McGill S, Grenier S, Bluhm M, Preuss R, Brown S, Russell C. Previous history of LBP with work loss is related to lingering deficits in biomechanical, physiological, personal, psychosocial and motor control characteristics. Ergonomics. 2003;46(7):731-46.
17. McKenzie DK, Bigland-Ritchie B, Gorman RB, Gandevia SC. Central and peripheral fatigue of human diaphragm and limb muscles assessed by twitch interpolation. J Physiol. 1992;454:643-56.
18. Mela P, Veltink PH, Huijing PA. The influence of stimulation frequency and ankle joint angle on the moment exerted by human dorsiflexor muscles. J Electromyogr Kinesiol. 2001;11(1):53-63.
19. Mizner RL, Petterson SC, Stevens JE, Vandenborne K, Snyder-Mackler L. Early quadriceps strength loss after total knee arthroplasty. The contributions of muscle atrophy and failure of voluntary muscle activation. J Bone Joint Surg Am. 2005;87(5):1047-53.
20. Nagesser AS, van der Laarse WJ, Elzinga G. Metabolic changes with fatigue in different types of single muscle fibres of Xenopus laevis. J Physiol. 1992;448:511-23.
21. Ng AV, Miller RG, Gelinas D, Kent-Braun JA. Functional relationships of central and peripheral muscle alterations in multiple sclerosis. Muscle Nerve. 2004;29(6):843-52.
22. Novy DM, Simmonds MJ, Olson SL, Lee CE, Jones SC. Physical performance: differences in men and women with and without low back pain. ArchPhys Med Rehabil. 1999;80(2):195-8.
23. Rumball JS, Lebrun CM, Di Ciacca SR, Orlando K. Rowing injuries. Sports Med. 2005;35(6):537-55.
24. Russ DW, Kent-Braun JA. Sex differences in human skeletal muscle fatigue are eliminated under ischemic conditions. J Appl Physiol. 2003;94(6):2414-22.
25. Scott WB, Lee SC, Johnston TE, Binkley J, Binder-Macleod SA. Contractile properties and the force-frequency relationship of the paralyzed human quadriceps femoris muscle. Phys Ther. 2006;86(6):788-99.
26. Scott WB, Oursler KK, Katzel LI, Ryan AS, Russ DW. Central activation, muscle performance, and physical function in men infected with human immunodeficiency virus. Muscle Nerve. 2007;36(3):374-83.
27. Stackhouse SK, Binder-Macleod SA, Lee SC. Voluntary muscle activation, contractile properties, and fatigability in children with and without cerebral palsy. Muscle Nerve. 2005;31(5):594-601.
28. Stackhouse SK, Dean JC, Lee SC, Binder-MacLeod SA. Measurement of central activation failure of the quadriceps femoris in healthy adults. Muscle Nerve. 2000;23(11):1706-12.
29. Stackhouse SK, Stevens JE, Johnson CD, Snyder-Mackler L, Binder-Macleod SA. Predictability of maximum voluntary isometric knee extension force from submaximal contractions in older adults. Muscle Nerve. 2003;27(1):40-5.
30. Stevens JE, Binder-Macleod S, Snyder-Mackler L. Characterization of the human quadriceps muscle in active elders. Arch Phys Med Rehabil. 2001;82(7):973-8.
31. Stevens JE, Stackhouse SK, Binder-Macleod SA, Snyder-Mackler L. Are voluntary muscle activation deficits in older adults meaningful? Muscle Nerve. 2003;27(1):99-101.
32. Suuden E, Ereline J, Gapeyeva H, Paasuke M. Low back muscle fatigue during Sorensen endurance test in patients with chronic low back pain: relationship between electromyographic spectral compression and anthropometric characteristics. Electromyogr Clin Neurophysiol. 2008;48(3-4):185-92.
33. Thomas JS, France CR, Lavender SA, Johnson MR. Effects of fear of movement on spine velocity and acceleration after recovery from low back pain. Spine. 2008;33(5):564-70.
34. Westerblad H, Bruton JD, Allen DG, Lannergren J. Functional significance of Ca2+ in long-lasting fatigue of skeletal muscle. Eur J Appl Physiol. 2000;83(2-3):166-74.


©2009The American College of Sports Medicine