THE CONCEPT that a reversal of antagonists could be

342 The Reversal of Antagonists Facilitates the Peak Rate of Tension Development David A. Gabriel, PhD, Jeffrey R. Basford, MD, PhD, Kai-Nan An, PhD ABSTRACT. Gabriel DA, Basford JR, An K-N. The reversal of antagonists facilitates the peak rate of tension development. Arch Phys Med Rehabil 2001;82:342-6. Objective: The present study was designed to test the effects of the reversal of antagonists on the peak rate of tension development (df/dt max ) of the elbow extensors. Design: Experimental, with matched controls. Setting: A biomechanics research laboratory. Participants: Twenty-six healthy women without a history of upper extremity injury or neurologic disorder, randomly assigned to experimental (n 13) or control (n 13) groups. Interventions: Two groups of healthy subjects followed identical exercise protocols, except that the control group performed maximal isometric contractions of the elbow extensors and the experimental group executed a maximal isometric elbow flexion contraction immediately before a maximal elbow extension contraction. Both groups performed 5 cycles of a 2-second contraction with 22-second rest periods between agonist muscle contractions and were evaluated at 4 test sessions spaced 2 weeks apart. Main Outcome Measures: All measurements were done with the shoulder and elbow at 90 of flexion in the sagittal plane to ensure reproducibility. A load cell was used to measure elbow extension moment and to calculate the peak rate of tension development (df/dt max ). Biceps and triceps brachii surface electromyographic activity was monitored concurrently. The electromyographic measures were mean spike (peak-to-peak) amplitude and mean spike frequency of the biceps and triceps brachii activity. Results: Intraclass df/dt max and electromyographic reliability was good (r.72) in both groups. Because biceps electromyographic measures were considerably less reliable (r.53), they were not included in our analysis. While df/dt max increased quadratically in both groups (p.05), the experimental group was on average 36.1Nm s 1 (63%) greater across sessions 2 to 4 (p.05). In contrast, triceps electromyographic activity did not differ significantly between groups (p.05). The means averaged across groups exhibited a quadratic increase from session 1 to session 4: 91 V or48% for mean spike amplitude (p.05) and 7Hz or 16% for mean spike frequency (p.05). Conclusions: The greater df/dt max for the experimental group was not associated with increased electromyographic From the Biomechanics Laboratory, Brock University, St. Catharines, Ont (Gabriel); and Departments of Physical Medicine and Rehabilitation (Basford) and Orthopedics (An), Mayo Clinic and Mayo Foundation, Rochester, MN. Accepted in revised form June 20, 2000. Supported by the Mayo Clinic/Mayo Foundation and the National Institutes of Health (training grant no. HD07447). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the author(s) is/are associated. Reprint requests to David A. Gabriel, PhD, Dept of Physical Education, Brock University, 500 Glenridge Ave, St. Catharines, Ont L2S 3A1, Canada, e-mail: dgabriel@arnie.pec.brocku.ca. 0003-9993/01/8203-6177$35.00/0 doi:10.1053/apmr.2001.21530 activity. The experimental group appeared to use the biomechanic properties of the pretensioned extensor muscle-tendon complex, rather than neurologic biasing, to accomplish its power gains. Key Words: Elbow; Electromyography; Isometric contraction; Rehabilitation; Shoulder. 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation THE CONCEPT that a reversal of antagonists could be used to enhance muscular performance was introduced by Kabat 1 in 1950. It was popularized by Knott and Voss 2 and is included in modern texts on therapeutic resistive exercise. 3 The basis of this idea is intriguing and quite simple: an isometrically contracting flexor muscles (eg, of the elbow) will help develop extensor strength in the antagonist muscles. In other words, an isometric flexion immediately before an extensor contraction facilitates extensor function through some sort of conditioning mechanism. The mechanism of this process is believed to be rooted in flexor Golgi tendon organ feedback produced during the conditioning flexor contraction, which is known to lower the firing threshold of extensor alpha motoneurons. 1 This proprioceptive feedback then is thought to combine with centrally generated commands to activate more motor units than would normally be recruited. 4 In an earlier study 5 comparing the reversal of antagonists that facilitate the elbow extensors with contractions of the extensors alone, we found no significant differences in strength or endurance. A review of the literature after that study 5 indicated that maximal inhibition by the Golgi tendon organs occurs within 1 second of the contraction. 6 Thus, there is a short time in which facilitation effects can be shown. In support for this theory, Grabiner 7 used an isometric conditioning contraction of the knee flexors to facilitate the knee extensors and found an increase in the peak rate of tension development (df/dt max ) and surface electromyographic activity, but not in maximal knee extension moment, which occurs later ( 1s) in the muscle contraction. Kabat s 1 view of Golgi tendon organs is oversimplified compared with what is known today. First, each organ is associated with only 3 to 25 muscle fibers, not to the whole muscle as originally conceived. 8 Furthermore, there are no more than 15 motor units in the muscle fibers attached to a single organ. 9,10 Different receptors, therefore, sense forces in different parts of the muscles with different temporal effects across the motor units. 9,10 Research on the efficacy of the reversal of antagonists may need to focus on changes in the rate of tension development and not on overall muscular strength. The purpose of the present study was to determine the effects of the reversal of antagonists on the peak rate of tension development and electromyographic activity during isometric elbow extension. METHODS Subjects The participants consisted of twenty-six sedentary righthanded women recruited from the general staff population at the Mayo Clinic. Men were excluded from the study because

REVERSAL OF ANTAGONISTS, Gabriel 343 they have trial and day error variances for isometric strength that are 2 to 4 times higher than that of women. 11 The exercise demands and experimental procedures were described to each subject, who then read and signed an informed consent document before participation. Weight training was not permitted during the study. Subjects were paid for their time. Apparatus and Testing Position The experimental apparatus and set-up has been detailed elsewhere 5 and is described here briefly. Subjects sat at a table designed to isolate the action of the elbow during flexion and extension contractions. A goniometer was used to position the shoulder and elbow of the arm being tested at 90 of flexion in the sagittal plane. The forearm was in neutral pronation and supination, and a wrist cuff was attached just below the styloid process of the wrist. The application of force was always perpendicular to the lever arm of the load cell. Adjustable support straps for the torso and shoulders were used to ensure stability and minimize extraneous movements. Measurement Schedule The 4 test sessions were scheduled at 2-week intervals. Subjects were ranked and randomly assigned by pairs into 2 groups and matched on the basis of the average df/dt max they developed during the 5 maximal isometric elbow extension trials on their first test session. Each subsequent session was conducted in an identical manner. The control group performed 5 cycles of a maximal 2-second isometric extension followed by a 24-second rest period. The experimental group followed a similar protocol modified to permit an isometric reversal of the antagonists: 5 cycles of a 2-second maximal isometric elbow flexion strength trial immediately followed by a 2-second maximal isometric elbow extension strength trial and a 22-second rest period. Instrumentation and Signal Processing Electromyographic activity was recorded with bipolar surface electrodes. Before electrode placement, the skin surface was shaved, lightly abraded, and cleansed with alcohol to limit the impedance at the skin-electrode interface to 5k for each subject. The electrodes were placed 3cm apart in line with the muscle belly and away from the motor points of the triceps long head and biceps short head. 12 The positions of the electrodes were marked with indelible ink to ensure the consistency of the placement across test days. The subjects were instructed to maintain the markings between sessions. Muscle activity was amplified up to 5000 times and band-passed between 10 and 300Hz. The JR3 a device was used to record elbow moment. All signals were digitized at 2016Hz using the DATAQ b computer-based oscillograph and data acquisition system and stored on a computer for later off-line processing. The force data were digitally low-passed (10Hz 3dB) with a zero-lag second-order Butterworth filter. The resulting curve was then differentiated using the finite-difference technique to obtain the peak rate of tension development (df/dt max ). The electromyographic activity produced during the tension development phase of the maximal isometric extension was used for analysis. The data window started when the rate of tension development was greater than 1% of df/dt max and ended when it fell below 20% of df/dt max after reaching its peak (fig 1). The 1% threshold avoided pretension on the load cell whereas the 80% criteria excluded the dramatic decline in df/dt and oscillation that occurs as maximal voluntary contraction is approached. 13 Mean spike (peak-to-peak) amplitude and mean spike frequency were calculated 14 for the triceps and biceps brachii. All data reduction was accomplished using a MAT- LAB c statistical package. Statistical Analysis The reliability of the criterion measures was first assessed using the intraclass correlation analysis of variance (ANOVA) model (r). 14 A split-plot factorial ANOVA model was then used to determine how the criterion measures responded to the reversal-of-antagonists technique. The between-block treatment was groups and the 2 within-block treatments were days and trials. 15 Trend analysis for means across test sessions was then conducted using orthogonal polynomials. 15 A probability level of.05 was defined as statistically significant. All statistical procedures were performed using SYSSTAT d software. RESULTS Tables 1 and 2 present the intraclass correlation ANOVA summary for each of the criterion measures for the control and experimental groups, respectively. The r values along with the trial and day error variance components suggest that most of the variability for df/dt max and triceps electromyographic activity was between subjects, not error variance associated with multiple trials or days. Thus, the reliability for these measures was good. In contrast, the electromyographic activity associ- Fig 1. A representative trial from a single subject. The rate of tension development (A), elbow extension moment (B), triceps brachii electromyographic activity (C), and biceps brachii electromyographic activity (D).

344 REVERSAL OF ANTAGONISTS, Gabriel Table 1: Intraclass Correlation ANOVA for the Control Group Component df df/dt max TMSA TMSF BMSA BMSF MS Subjects 12 9873.633 395766.464 1113.059 50960.062 1561.116 MS Days within Subjects 39 1672.336 101048.531 222.968 105011.990 732.873 MS Within Cells 208 ( e1 Trials) 158.211 20344.963 72.431 5005.391 160.885 e1 Days 302.825 16140.714 30.107 20001.320 114.398 e1 True 410.065 14825.897 44.505 2702.56 41.412 r.83.75.80.34.53 Abbreviations: TMSA, triceps mean spike amplitude; TMSF, triceps mean spike frequency; BMSA, biceps mean spike amplitude; BMSF, biceps mean spike frequency. ated with biceps coactivity was not reliable and was omitted from further analysis. Figure 2 shows the means and standard deviations (SDs) for df/dt max for the control and experimental groups. The means for both groups were nearly identical on session 1 when they performed maximal isometric elbow extension strength trials only. Thus, the 2 groups were matched well. A marked increase occurred between the first and second session, and then the means leveled off after the second session. This resulted in significant linear and quadratic trend components accounting for over 99% of the total variance in means across sessions for both groups (p.05). However, the between-groups interaction terms were significant (p.05) for both the linear and quadratic components: the experimental group was on average 36.1Nm s 1 (63%) greater than the control group after the first session. Statistical analysis of triceps mean spike amplitude failed to show any significant differences between groups (p.05). As might be expected, triceps mean spike amplitude followed the same pattern of change as df/dt max across sessions (fig 3). Together, both groups had an average increase in triceps mean spike amplitude of 91 V (48%) from the first to second session (p.05). We found no further increase after the second session. Both groups, therefore showed significant linear and quadratic trend components accounting for over 99% of the total variance in means across sessions (p.05). The pattern of change in triceps mean spike frequency was slightly different from the other 2 criterion measures. The means increased from session 1 to session 3 but plateaued between sessions 3 and 4 (fig 4). This adaptation pattern also resulted in significant linear and quadratic trend components accounting for over 99% of the total variance in means across sessions (7Hz or 16%; p.05). The 2 groups, however, never differed by more than 3Hz (p.05). DISCUSSION The present study was designed to test the effects of the reversal of antagonists on the peak rate of tension development (df/dt max ) of the elbow extensors. A 2-second contraction of the elbow flexors immediately before the extensors was found to increase df/dt max without a concomitant rise in either extensor mean spike amplitude or spike frequency. Intraclass correlation ANOVA showed that the criterion measures were reliable, except for antagonist coactivity. However, it is not unusual for agonist activity to have good reliability whereas antagonist coactivity shows considerable error variance. 16 The experimental and control groups increased in df/dt max from session 1 to session 3 while there was a plateau between sessions 3 and 4. It is generally agreed that increases in muscular performance after a limited number of contractions can be attributed to motor learning. 17 The plateau in df/dt max before the end of the measurement schedule supports this hypothesis. 18 It is important to note that the more complex contraction pattern required by the reversal of antagonists did not interfere with motor learning. 5 Although both groups increased in df/dt max across sessions, the experimental group averaged a 63% greater change than the control group without a significant difference in electromyographic activity between the groups. This lack of difference is not an artifact of limited statistical power because the means and SDs for electromyographic activity between the 2 groups clearly overlapped. 19 Within the limitations of the study technique (surface electromyographic measurement), the results show a lack of evidence for neural factors. There are 3 well-documented phenomena that are pertinent here to this issue. 20-22 One possibility involves the use of elastic energy stored in the muscle and tendons. 20 During the conditioning contraction of the elbow flexors, the elbow extensors underwent excitation and contraction coupling wherein the viscoelastic components of the muscles were placed in tension. Elastic recoil of the extensor muscle-tendon complex might then be used to augment the rate of tension development during maximal isometric contractions of the elbow extensors in the following phase. However, because there is limited shortening in isometric actions of the muscle, it is unlikely that elastic recoil plays a significant role. In the absence of elastic recoil, Table 2: Intraclass Correlation ANOVA for the Experimental Group Component df df/dt max TMSA TMSF BMSA BMSF MS Subjects 12 26703.939 2932486.799 1433.239 454918.728 1135.195 MS Days within Subjects 39 7496.023 764552.398 330.132 232744.072 641.065 MS Within Cells 208 ( e1 Trials) 396.013 19753.157 108.788 13936.363 150.138 e1 Days 1420.002 148959.848 44.269 43761.542 98.185 e1 True 960.396 108396.720 55.155 11108.733 24.706 r.72.74.77.49.44

REVERSAL OF ANTAGONISTS, Gabriel 345 Fig 2. The means and SDs (vertical bars) for the peak rate of tension development (df/dt max ) for the experimental and control groups. Fig 3. The means and SDs (vertical bars) for triceps electromyographic mean spike (peak-to-peak) amplitude for the experimental and control groups. Fig 4. The means and SDs (vertical bars) for triceps electromyographic mean spike frequency for the experimental and control groups. a second benefit of a taut muscle-tendon complex is on contraction dynamics. The slack taken out of extensor muscletendon complex might permit rapid tension development and for its transmission to the bone during the following extensor contraction. 21 Third, a limited stretch of the extensors may take advantage of the length-velocity relationship. 22 The length of the extensors during the flexion conditioning contraction may be closer to resting muscle length (L 0 ), which has the highest velocity of shortening for a given load. The velocity of shortening for a given load is less when the contraction is initiated at a muscle length shorter than L 0, as might be the case for the extensors without a conditioning contraction of the flexors. We previously showed no difference between groups with respect to maximum strength, 5 yet our present subjects differed on the basis of df/dt max. This observation corroborates the earlier results of Grabiner 7 and supports our contention at the outset that facilitation occurs early during the tension development phase of the contraction. At present, the exact mechanism remains unclear. Future work will assess changes in H-reflex amplitude in the agonist after antagonist contraction at 2 different joint angles. If muscle length is a factor, then the response to a conditioning contraction of the antagonist should depend on joint angle during isometric contractions. The increase in agonist df/dt after a conditioning contraction of the antagonist should be less for a joint angle that lengthens the agonist muscle-tendon complex than for a joint angle that shortens it. At the same time, there should be no change in either electromyographic activity or H-reflex amplitude of the agonist. CONCLUSIONS The reversal-of-antagonists technique was superior to agonist-only exercise for increasing the peak rate of tension development. There was a lack of evidence for neural factors. The complex contraction pattern required by the reversal of antagonists did not interfere with increases in muscle performance associated with motor learning. References 1. Kabat H. Studies on neuromuscular dysfunction: XV. The role of central facilitation in the restoration of motor function in paralysis. Arch Phys Med Rehabil 1950;33:521-33. 2. Knott M, Voss DE. Proprioceptive neuromuscular facilitation. 2nd ed. New York: Harper & Row; 1968. 3. Hanson C. Proprioceptive neuromuscular facilitation. In: Hall CM, Brody LT, editors. Therapeutic exercise: moving toward function. Philadelphia: Lippincott Williams & Wilkins; 1999. p 233-51. 4. Harris FA. Facilitation techniques and technological adjuncts in therapeutic exercise. In: Basmajian JV, editor. Therapeutic exercise. 4th ed. Baltimore: Williams & Wilkins; 1984. p 110-78. 5. Gabriel DA, Basford J, An K-A. Effects of the reversal of antagonists upon isometric elbow extension strength and endurance. Arch Phys Med Rehabil 1997;78:1191-5. 6. Moore MA, Kukulka CG. Depression of Hoffmann reflexes following voluntary contraction and implications for proprioceptive neuromuscular facilitation. Phys Ther 1991;71:321-33. 7. Grabiner MD. Maximum rate of force development is increased by antagonist condition contraction. J Appl Physiol 1994;77:807-11.

346 REVERSAL OF ANTAGONISTS, Gabriel 8. Stuart DG, Mosher CG, Gerlach RL, Reinking RM. Mechanical arrangement and transducing properties of Golgi tendon organs. Exp Brain Res 1972;14:274-92. 9. Appenteng K, Prochazka A. Tendon organ firing during active muscle lengthening in awake, normally behaving cats. J Physiol (Lond) 1984;353:81-92. 10. Gregory JE. Relations between identified tendon organs and motor units in the medial gastrocnemius muscle of the cat. Exp Brain Res 1990;81:602-8. 11. Kroll WP. Test reliability and errors of measurement at several levels of absolute isometric strength. Res Q Exerc Sport 1970;41: 155-63. 12. Delagi EF, Perotto AP. Anatomic guide for the electromyographer. 2nd ed. Springfield (IL): CC Thomas; 1980. 13. Kamen G. The acquisition of maximal isometric plantar flexor strength: a force-time curve analysis. J Mot Behav 1983;15:63-73. 14. Gabriel DA. Reliability of SEMG spike parameters during concentric contractions. Electromyogr Clin Neurophysiol 2001;40: 1-8. 15. Kirk RE. Experimental design: procedures for the behavioral sciences. Belmont (CA): Brooks-Cole; 1968. 16. Gabriel DA, Kroll WP. Isometric successive induction resistance exercise. Clin Kinesiol 1991;45:30-7. 17. Kramer WJ, Fleck SJ, Evans WJ. Strength and power training: physiological mechanisms of adaptation. In: Holloszy JO, editor. Exercise and sport sciences reviews. Vol 24. Baltimore: Williams & Wilkins; 1996. p 363-97. 18. Kroll WP. Analysis of local muscular fatigue patterns. Res Q Exerc Sport 1981;52:523-39. 19. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale (NJ): Lawrence Erlbaum; 1988. 20. Proske U, Morgan DL. Tendon stiffness: methods of measurement and significance for the control of movement. A review. J Biomech 1987;20:75-82. 21. Zierler KL. Some aspects of the biophysics of muscle. In: Bourne GH, editor. The structure and function of muscle. Vol 3. New York: Academic Pr; 1973. p 117-83. 22. Cavagna GA. Storage and utilization of elastic energy in skeletal muscle. In: Hutton RS, editor. Exercise and sport sciences reviews. Vol 5. Santa Barbara (CA): Journal Publ Affiliates; 1977. p 89-129. Suppliers a. JR3 Inc, 22 Harter Ave, Woodland, CA 95776. b. DATAQ Instruments, Inc, 150 Springside Dr, Ste B220, Akron, OH 44333. c. The MathWorks, Inc, 24 Prime Park Way, Natick, MA 01760. d. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.