An Electromyographic Study of Unresisted Trunk Rotation With Normal Velocity Among Healthy Subjects [Biomechanic]

Transcription

1 Page 1 of 22 Lippincott-Raven Publishers. Volume 21(13) 1 July 1996 pp An Electromyographic Study of Unresisted Trunk Rotation With Normal Velocity Among Healthy Subjects [Biomechanic] Kumar, Shrawan DSc, PhD; Narayan, Yogesh BS; Zedka, Milan MD From the Department of Physical Therapy, University of Alberta, Edmonton, Alberta, Canada. Supported by the Medical Research Council of Canada, Ottawa. Acknowledgment date: May 25, First revision date: December 19, Second revision date: May 9, Third revision date: August 3, Acceptance date: November 15, Device status category: 1. Address reprint requests to: Shrawan Kumar, DSc, PhD; 3-75 Corbett Hall; Department of Physical Therapy; University of Alberta; Edmonton, Alberta; T6G 2G4 Canada Outline Abstract [black small square] Materials and Methods [black small square] Results Phasic Relationship Slopes Electromyographic Magnitude [black small square] Discussion References Graphics Table 1 Figure 1 Figure 2 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Equation 1A

2 Page 2 of 22 Figure 8 Figure 9 Figure 10 Figure 11 Abstract Study Design: An axial rotation tester was designed and fabricated for the study. This allowed stabilization of seated subjects (hip down) and coupling of shoulders, permitting axial rotation and coupled lateral flexion. Using this device, a flexion-extension free axial rotation was executed for studying its characteristics. Objectives: To determine the mechanism of initiation, sustenance, and execution of axial rotation. This was planned to be done by determining the phasic relationship of various torso muscles in the initiation, execution, and termination of axial rotation. Another objective was to determine the total and relative contribution of torso muscles in axial rotation and the small segments of these activities. Summary of Background Data: There only are a few studies conducted on axial rotation. Generally, these have investigated isometric maximal voluntary contraction in neutral or prerotated postures. The two studies that have reported isokinetic axial rotation have investigated maximal efforts. No study in literature has reported initiation, termination, and execution of unresisted normal velocity axial rotation. Methods: Fifty healthy young subjects executed a full cycle of axial rotation, starting from neutral position to their extreme left, continuing to their extreme right, and finally moving to the neutral posture in one smooth motion without stopping anywhere. The electromyographic results of external obliques, internal obliques, rectus abdominis, pectoralis major, erectores spinae at T 10 and L 3, and latissimus dorsi were measured bilaterally simultaneously during this trunk rotation. The timing and relative magnitude analyses were done to determine the global and individual muscle contributions in axial rotation. The correlation between electromyographic and angular displacement, and nonlinear curve fitting regression analyses were performed to decipher individual muscles behavior. Results: The pattern of muscle activation was variable. However, contralateral external obliques, ipsilateral erector spinae, and latissimus dorsi became active before other muscles. These were agonists and the others were antagonists or stabilizers. The agonists contributed 65% of the total electromyographic output, whereas antagonists and stabilizers contributed 35%. The muscle activities during onset and offset periods were biphasic with significantly different slopes. Conclusions: It was concluded that the axial rotation is achieved through the activities of agonists, and return to neutral position is because of elastic recoil controlled by agonistic muscles. A range of approximately on either side of the anatomical midsagittal plane involves little muscle effort, but beyond this region, the osteoligamentous structures become stiff and require increasing effort to execute axial rotation. Information on mechanics of rotation as well as mechanism of development of rotary torque is sparse. The phasic relationship as well as relative quantitative activity of different muscles is unclear and incomplete. Although it is known that both abdominal and dorsal muscles are involved in the development of axial torque, its initiation, relative contribution, control and stabilization, and relaxation remain to be studied precisely and quantitatively. An antagonistic activity of deep lumbar trunk muscles during axial rotation has been reported.2 Occasional activity in the ipsilateral multifidi and rotatores during axial rotation also has been

3 Page 3 of 22 reported.3,11 Others have found high levels of activity on the contralateral side when positions of lateral bend and rotation were combined with an external load.1 Pope et al 12 showed that antagonistic activity was much higher than reported in any previous work. They found little difference in activity between the left and right sides of the many muscle pairs. However, erector spinae did not show significant difference between ipsilateral and contralateral sides. The greatest magnitude of electromyographic activity was reported in the erector spinae muscles. The effect of 30 prerotation in the direction of torque development was found to decrease marginally the maximal torque, whereas prerotation in the opposite direction increased the developed torque.13 Postural asymmetry is reported to influence torque generation in isometric and isokinetic activities.6 Electromyographic (EMG) activity of abdominal (i.e., external and internal obliques and rectus abdominis) and spinal (i.e., latissimus dorsi, upper and lower erector spinae) muscles in isometric and isokinetic maximal voluntary contraction was reported by McGill.10 McGill also reported less than maximal muscle activity in torsional maximal voluntary contraction when normalization of EMG activity was done based on nontorsional maximal EMG values (ranging between 22-74%). In an isokinetic effort, he reported a dominant role played by the latissimus dorsi and external oblique muscles. Upper erector spinae also was found linked with generation of axial torque. Marras and Mirka 9 reported an experiment studying EMG and intraabdominal pressure during trunk loading in asymmetric motion. They reported significant response of trunk muscles to the trunk angular velocity, trunk torque level, and trunk posture. They reported interrelated trends in EMG activity of trunk muscles. In this study, the EMG was filtered at 80 Hz, a value close to the median frequency of most trunk muscles. Therefore, it is not clear if the responses of these muscles truly are representative of the muscle response. Furthermore, single values at selected positions were reported; the mechanism of initiation and sustenance of rotation was not investigated. Kumar 4 reported for the first time the phasic relationship between the dorsal and abdominal muscles in axial rotation. However, the current knowledge remains incomplete in delineation of the role of abdominal and dorsal muscles in initiation, magnitude of contribution, phasic relationship, and also the extent of synergism and antagonism of different spinal muscles during axial rotation. To answer some of these questions, a study was launched where continuous EMG of dorsal and abdominal muscles was monitored through complete cycles of axial rotation. [black small square] Materials and Methods Subjects. Fifty young subjects, all of whom were University of Alberta students, volunteered for the experiment. Of these, 27 were men with a mean age of 22 years, mean height of 177 cm, and mean weight of 72.3 kg. The remaining 23 subjects were women with mean age of 22 years, mean height of cm, and mean weight of 58.5 kg. The details of the relevant demographic characteristics are listed in Table 1. All subjects were healthy with no musculoskeletal disorders and without history of back pain within the last year. The exclusion criterion was scoliosis, spinal, abdominal surgery, or pregnancy. Such volunteering subjects were provided with the purpose and procedures of the project. Consenting subjects signed the informed consent form before proceeding onto the experiment. Table 1. Demographic Characteristics of the Experimental Sample These subjects were weighed and measured for their height, seat-to-shoulder height, length of trunk and head, intershoulder (left-to-right-acromion processes) width, and the width of the intershoulder

4 Page 4 of 22 harness. Their age was recorded. These subjects then were applied 14 pairs of disposable pregelled surface electrodes (HP ) at an interelectrode distance of 2.5 cm after suitable preparation of the skin with an alcohol-acetone mixture. These electrodes were placed on erectores spinae level with spinous processes of T 10 and L 3 vertebrae bilaterally. Surface electrodes also were applied to left and right latissimus dorsi. On the ventral side, these surface electrodes were applied bilaterally to the pectoralis major, the rectus abdominis, the external oblique, and the internal obliques. A ground electrode was applied to anterosuperior iliac spine. The subjects were wired to an isolated preamplifier system to provide onsite amplification. Such prepared subjects were seated in the chair of the axial rotation tester. The seat was adjusted for the height to have 90 at the knee. The seat also was aligned with the axial rotation harness, which was lowered on the subject's shoulders and fastened. The subjects were stabilized in this seated position hip down using four Velcro straps at hip, distal thigh, proximal shin, and ankle. The subjects then were asked to rotate to their extreme left and right to ensure that there was no restriction to motion. If such a restriction was encountered, the chair was readjusted until restriction-free floating axial rotation was obtained. Tasks. These subjects were asked to assume their neutral position and sit still for a few seconds. From this resting position, the subjects were asked to rotate on command go to their extreme left rotation and without stopping to reverse the direction of rotation going to their extreme right, finally returning to the initial neutral position without stopping anywhere in between. This total excursion was considered one cycle of axial rotation. This cycle was divided into four phases: 1) neutral (0 ) to left, 2) left to neutral (0 ), 3) neutral (0 ) to right, and 4) right to neutral (0 ). Equipment. The experiment was conducted using an axial rotation tester for posture restriction and motion execution. The EMG was recorded through EMG preamplifiers and amplifiers, and the central controller computer (486) with a Metra Byte DAS 20 A to D board (Metra Byte Corp., Taunton, Massachusetts). These are described briefly below. Axial Rotation Tester. The axial rotation tester used was after Kumar 5 Briefly, it consisted of a cage of metal frame mounted on a metal baseplate. A backrest sawed-off chair with four sets of adjustable belts was mounted so that on the baseplate, it could be slid back and forth and raised and lowered by a jack underneath. Mounted on the top bars was a spring-loaded metal hollow cylinder in a sleeve with a locking screw. At the lower end of this cylinder, a circular metal disc with a groove in its rim was welded. To this circular disc, an upholstered metal shoulder harness on the two arms of an inverted Y was mounted (Figure 1). The arms of the harness were adjustable for width as well as opening. The rotating cylinder was coupled with a high-precision potentiometer through a set of gears to measure rotation. The potentiometer was linear within 0.5% for a range of 180.

5 Page 5 of 22 Figure 1. Schematic diagram of an axial rotation tester. Electromyographic Setup. The EMG setup consisted of disposable pregelled surface electrodes (HP ) and the shielded short leads feeding the preamplifiers. These fully isolated preamplifiers with two stages of gain (10 or 100) were built to specification with low and high cutoff filters at 30 Hz and 300 Hz, respectively. The preamplifiers fed to the 16-channel amplifier through shielded cable and input connectors. The amplifier had eight levels of gains, a 60-Hz notch filter, and output options for raw EMG signals or their root mean square derivative. The EMG equipment was built by Woolly Electronics (Ann

6 Page 6 of 22 Arbor, MI). The preamplifier and amplifier combination was calibrated for a linear response within a range of 100 µv-5 mv. Controller and A to D Board. The outputs of the axial rotation tester potentiometer and the EMG amplifiers were fed to a Metra Byte DAS 20 A to D board. This A to D board was capable of sampling at up to 100 khz. Such converted digital signals were stored in the hard disc of a 486 with a tape backup (Colorado Memory Systems Inc.). A specialized software was developed for this project that enabled data acquisition and analysis. This data acquisition modular software was an expansion of the software published by Kumar and Garand.6 Data Acquisition. For data acquisition prepared, seated and stabilized subjects were issued the command go approximately 1 second after beginning to sample. The sampling was done for a period of 10 seconds. At the command, the subjects axially rotated to the end of their range of left rotation and reversed the direction to the opposite direction to the end of their range of motion to the right, and finally returned to the neutral position. The entire cycle was performed in one smooth and continuous movement. The 14 EMG channels were set on identical gain of 10,000 and were band passed between Hz. These EMG signals were sampled at 50 Hz after converting them into their root mean square. A time constant of 20 msec was chosen. Analysis. The collected data were loaded in the computer, and windowing was done for analysis (Figure 2A). When the trunk rotation reached a threshold level of 2% of the maximal angular rotation at the beginning and the end of the cycle, the computer software automatically drew a vertical line, marking the start and the end of the cycle. A smoothing routine was used to smooth the signals to reliably interpret the pattern. Using the start position as the point of onset, all other timings were measured with reference to it. The times of maximum angular rotation, duration of the task cycle, time of anticipation, onset, offset, and relaxation for all 14 EMG channels were determined. For this determination, each channel of EMG was plotted with the angular displacement and task cycle individually (Figure 2B). The relevant sections were amplified in case of any ambiguity. Using a cursor, the points of beginning anticipation, a clear onset, the point of offset, and the point of relaxation were marked and stored in the memory. Using the cycle start time from the initial activity window, the computer calculated the specific times of these events. The software automatically joined these points to delineate the segment of anticipation, onset, offset, and relaxation. From these points, the software calculated the slopes of these segments. The software also identified the maximum EMG score of all channels, the time of the maximum EMG, the angular displacement at maximum EMG, the average EMG, the average EMG per second, the EMG area, the percent contribution by each channel individually, and the total EMG in the activity of unresisted normal velocity axial rotation. Finally, the software divided the chosen activity in segments of 10% of the task cycle and measured the magnitude of EMG, the total EMG at that percent of cycle, and individual channels as percent of this total EMG. A similar analysis of EMG area of each channel was done automatically by the analysis software.

7 Page 7 of 22 Figure 2. A, Windowing and marking the task cycle by angular displacement trace. B, Superimposition of the electromyographic trace of the right external oblique muscle on the angular displacement trace and marking the traces. A statistical analysis of these EMG parameters was carried out to calculate descriptive statistics, analysis of variance (significance level = 0.05), and the Pearson's product-moment correlation between EMG and angular displacement. A curve fitting to the data also was done to determine the regression

8 Page 8 of 22 equation representing the variable behavior. Limitations of the Electromyographic Technique. The use of a common gain and normalization based on maximal voluntary contraction may introduce errors in the data where absolute magnitude or comparison of magnitude across channels may be done. Because of varying distances of different muscles from the surface, the EMG pickup is affected differentially in different muscles about their force output. Further, as the dynamic contraction proceeds, changes in pickup volume may occur unrelated to muscle power output. The forgoing are considered to be the limitations of the study. [black small square] Results The traces of the unresisted normal velocity axial rotation indicated a distinct and different phasic and quantitative relationship among the torso muscles. The starting neutral position frequently was different from 0. The mean angular deviation from the neutral position among men was between 1.73 to the right and 3.3 to the left at the start and finish, respectively. These mean values of the angular deviations for women were between 2.9 to the left and 4 to the right at the start and finish, respectively. The mean total time taken in the complete cycle by men was 5.33 seconds and by women was 5.4 seconds. When the total cycle was broken into its four phases (0 to left, left to 0, 0 to right, and right to 0 ), the angular rotation varied between for men and 70.1 and 71.3 for women for full-range deviations to the left and right, respectively. Each of these phases took between seconds (Table 2) to complete. Table 2. Cycle Time Angular Displacement and the Phase Time Phasic Relationship The phasic relationship between different EMG channels about the angular displacement as seen on the traces was consistent in trend, but variable in magnitude. The beginning of onset, accelerated response, offset, and relaxation for different muscles occurred at significantly different times (P<0.001). The analyses of variance showed that these variations were unique for the muscles but were not affected by the gender. Furthermore, there was no interaction between the muscle and the gender, implying that the two genders did not use different strategies in carrying out axial rotation. Analyses of variance carried

9 Page 9 of 22 out on timing for each muscle individually showed no significant differences between the subjects, except in the left pectoralis (P<0.02). In addition to a statistically significant difference in the timing relationships of different muscles with the events, there was a considerable variation between subjects in the recruitment pattern. However, there was no statistically significant effect due to subjects. The mean values of the timings of onset, accelerated response, offset, and relaxation for men and women are presented in Tables 3 and 4. The consistent pattern was that the contralateral external oblique and the ipsilateral erectores spinae of the thoracic region were the major initiators of the axial rotation. These muscles did receive help in sustenance and execution of the axial rotation by the ipsilateral latissimus dorsi, internal oblique, and the lumbar erectores spinae somewhat later in the initial stages. All other muscles became active in advanced stages of axial rotation. However, for the offset and relaxation, all muscles in both genders decreased their firing significantly much before the angular deviation slowed down or stopped (Tables 3 and 4). Table 3. Mean Onset and Decay Times of Different Muscles in Relation to Angular Displacement Among Males

10 Page 10 of 22 Table 4. Mean Onset and Decay Times of Different Muscles in Relation to Angular Displacement Among Females Slopes In response to the axial rotation, the variable magnitude contribution occurred not only with different phasic relationships but also with different rates of rise. The analyses of variance showed that the slopes of different muscles were significantly different from each other (P<0.001). Furthermore, the two genders evoked significantly different slopes of the rise and decay of different muscles (P<0.001). Such was the case for onset, accelerated response, offset, and relaxation. Among the specific muscles, the right external obliques, left erectores spinae at T 10, and left latissimus dorsi had significantly different (steeper) slopes in the onset segment (P<0.05). The left erectores spinae at L 3, and at T 10, left latissimus dorsi, and right external obliques had significantly different (higher) slopes during the offset phase of the axial rotation activity. However, all other muscles and the angular displacement did not vary significantly in their slopes. The slopes of the onsets and the accelerated responses were significantly different from each other (P<0.001). The point of inflection from the slope of onset to the slope of accelerated response was variable for different muscles but occurred within the envelope of trunk rotation. Electromyographic Magnitude

11 Page 11 of 22 The recorded peak EMG activity and the EMG area for different muscles were significantly different from each other (P<0.001). Also, the peak EMG in different phases of axial rotation for different muscles was significantly different from each other (P<0.001). The latter was clearly reinforced by a significant two-way interaction between the muscle and the phase of the activity (P<0.001). The peak and the area of EMG of each of the 14 muscles in four phases also were significantly different from each other. The peak EMG and the EMG area recorded among men were significantly different from those among women (P<0.001). The individual values of the peak EMG, the time of the peak EMG, the angular displacement at the peak EMG, the average EMG magnitude, the area of EMG of individual muscles, and the area of muscles as percentage of the total EMG output for each of the phases of the axial rotation for men are presented in Tables 5 through 8. The female subjects had a similar pattern. The Pearson's product-moment correlation analysis with 95% confidence interval showed the following data. The peak EMG was highly correlated with average EMG (range, ; P<0.001). Similarly, the peak EMG was strongly correlated with the EMG area (range, ; P<0.001). The average EMG and the EMG area also were mutually correlated (range, ; P<0.001). Table 5. Mean Electromyographic Magnitude Parameters for Leftward Axial Rotation Among All Males Table 6. Mean Electromyographic Magnitude Parameters During Return From Left to Neutral Position Among All Males

12 Page 12 of 22 Table 7. Mean Electromyographic Magnitude Parameters for Rightward Rotation From Neutral Position Among All Males Table 8. Mean Electromyographic Parameters During Return From Right to Neutral Position Among All Males The total muscle output as measured in units of millivolts per second was similar to that for Phases 1 and 3 and for Phases 2 and 4 for men and women alike. Phases 1 and 3 generated twice as much EMG activity as did Phases 2 and 4 in both sexes. Phases 1 and 3 generated mv/sec and mv/sec among men, respectively. Phases 2 and 4, conversely, generated mv/sec and mv/sec, respectively, among men. The women generated mv/sec and mv/sec for Phases 1 and 3, and mv/sec and mv/sec for Phases 2 and 4, respectively. The distribution of the EMG activity over the different segments of the cycle as percent of total EMG is shown in Figure 3 for the male sample. The responses of women were similar to those of men. In the first and third phases of the axial rotation task, most of the EMG activity took place after the first 40% of task cycle. Whereas in Phases 2 and 4, there was a precipitous decline of the EMG activity in the first 40% of the task cycle with little activity beyond this stage.

13 Page 13 of 22 Figure 3. The electromyographic output of the ventral muscles during 0 to left axial rotation among men at 10% interval of the task labeled as percentile. However, within the 10 task-cycle segments in Phases 1 and 3, the contralateral external oblique began the activity and continued to increase until the end. In the first segment, it constituted 12.8%, and in the 10th segment, it constituted 21.1% of the total EMG for these respective segments. The initial rate of increase was more rapid and by the fourth segment, it had stabilized at 20% of total segment output. A somewhat similar pattern of activity was found in the ipsilateral latissimus dorsi, increasing from 12% in the first segment to 24% in the 10th segment, having reached 20% in the fifth segment and increasing slowly thereafter. These values, as percent of the total for the same muscle, undergo steady increase with rate increasing further toward the extreme of the range of motion. The total increase was more than 10- fold. The ipsilateral erector spinae had an opposite pattern. At L 3 level, it started with 15% in the first segment, declining gradually and smoothly to 7% in the 10th segment. At T 10 level the erector spinae started with 25% of the first segment and declined to 10% of the 10th segment. The initial decline up to the fourth segment was considerably steeper leveling off thereafter. A similar magnitude and pattern were found among women for these phases. In both genders, however, when one considered the relative contribution of the muscle within the segment as proportion of the total EMG activity of that muscle in the phase, these activities increased up to threefold. The pattern for Phases 2 and 4 among the men and women both were somewhat like mirror images of Phases 1 and 3, except the rate of change was much faster in the first four segments. When each of the four phases of axial rotation was divided into 10% interval of the respective task cycle and the EMG magnitude of each of the muscles plotted at these percent levels of the task cycles, the pattern of the EMG contribution became clearer. These percentile plots for the male experimental sample are presented in Figures 3 through 6. The response of the female sample was similar to the males in pattern, though they were different in exact magnitudes. However, among both genders in the first phase (0 to left), the maximal EMG activities were recorded at 80% to 90% of the task cycle where angular displacement was almost maximal, but the motion was still progressing toward 100% of the task cycle.

14 Page 14 of 22 Upon stopping at 100% cycle, there was a drop in the magnitude of all muscles (Figures 3, 4). The highest contributing muscles were right external oblique, left latissimus dorsi, and the erector spinae at L 3 and T 10 levels. While returning to the neutral posture, the maximal EMG was recorded at 0% of the cycle at the initiation of the activity. In this phase, the highest levels of activities again were recorded in the right external obliques, left latissimus dorsi, and the erectores spinae at T 10 and L 3 levels. The magnitude of the EMG dropped precipitously within the first 20% of the task cycle (Figures 5, 6). The task percentile scores for Phases 3 and 4 (0 to right, right to 0 ) were mirror images of Phases 1 and 2, except the muscles were on the other side of the body. The total EMG output always was greater when the rotation was from 0 to left or 0 to right (Phases 1 and 3) (Figure 7). During the return toward the neutral posture from the left and right rotations (Phases 2 and 4), the total EMG output was significantly less, and the decline in the muscle activity was rapid (Figures 8, 9). A precipitous decline took place in the first 20% of the cycle, and by 40% of the cycle, the total EMG output reached nearly resting level. Figure 4. The electromyographic output of the dorsal muscles during 0 to left axial rotation among men at 10% interval of the task labeled as percentile.

15 Page 15 of 22 Figure 5. The electromyographic output of the ventral muscles during left to 0 axial rotation among men at 10% interval of the task labeled as percentile.

16 Page 16 of 22 Figure 6. The electromyographic output of the dorsal muscles during left to 0 axial rotation among men at 10% interval of the task labeled as percentile. Figure 7. The electromyographic area of percentile segment as percent of total electromyographic during axial rotation among men.

17 Page 17 of 22 Figure 8. The combined total electromyographic output of all muscles in relation to the angular displacement during axial rotation from 0 to left among men at 10% interval of the task labeled as percentile.

18 Page 18 of 22 Figure 9. The combined total electromyographic output of all muscles in relation to the angular displacement during axial rotation from left to 0 among men at 10% interval of the task labeled as percentile. The correlation between total EMG and the angular displacement for Phases 1 and 3 ranged between for both men and women (P<0.001). The correlation coefficients for Phases 2 and 4 for both sexes were significantly lower, ranging between (P<0.02). A nonlinear regression analysis performed for the EMG output for various task percentiles was significant (P<0.01) and faithfully predicted EMG of each muscle over the entire task (Figures 10, 11). The equation that fitted the data is given below: Equation Equation 1A

19 Page 19 of 22 Figure 10. Nonlinear regression curve fitted to the electromyographic activities of different muscles for neutral to left rotation among men at 10% interval of the task labeled as percentile.

20 Page 20 of 22 Figure 11. Nonlinear regression curve fitted to the electromyographic activities of different muscles for left to neutral rotation among men. where B 0 = constant; B 1 = magnitude of the curve; B 2 = mean of percentile EMG; B 3 = standard deviation; and z = percentile - B 2 /B 3. [black small square] Discussion During the experiment, the importance of achieving the initial neutral position and executing the axial rotation with a smooth normal velocity was emphasized to all subjects. It is interesting to observe that

21 Page 21 of 22 rarely, subjects assumed an objective 0 initial neutral position or even returned to their original starting position. The mean range of deviation in the sample was found to be approximately 5.0 among men and 3.3 among women. it is worthy of note that these subjects were stabilized hip down and by shoulder harness. Considerable proprioceptive input through these contacts as well as visual feedback was available. Though the concept of a neutral zone in spinal articulation was proposed recently in vitro,14 its existence in vivo was first reported by Kumar and Panjabi 7 and subsequently by Kumar and Panjabi.8 The presence of such a neutral zone indicates a region analogous to the toe region of the stress strain curves for collagenous structures. In fact, the pattern of EMG activity of all active muscles during the phase of initiation as well as decay was biphasic. During the first segment of recruitment, a gradually increasing low level EMG was recorded. On reaching an angular rotation of approximately 10-15, the EMG response (slope) increased rapidly. Clearly this implies that the cumulative joint play at intersegmental levels allowed by the geometric design of these joints, and the ligamentous restraint system of these joints had become taut and begun to offer resistance to the rotational efforts at this point. During the return phase of the activity, the EMG output, although high in the beginning, dropped precipitously within the first 20% of the task cycle, indicating an elastic recoil. The total EMG output during return was significantly lower than that of during active axial rotation. Furthermore, during the return phase, the muscles were silenced completely up to 0.75 second before the completion of the cycle. During this phase, the agonists of the first phase also were active controlling the rate of return. Finally, during active axial rotation as well as during return to the neutral, the angular displacement was a smooth and gradual curve, even though the EMG response was biphasic. These observations clearly indicate to an initial lax region followed by a tense and taut range. A total axial rotation on either side of between among men and women both indicated that there was no gender difference in the range of axial rotational motion. The task cycle between seconds for men and women indicates that the subjects naturally moved at a comparable rotational velocity of approximately between per second. Therefore, the pattern of the responses obtained from the two genders is reliable as well as comparable. As has been previously reported 4,9,10 the external obliques were active during axial rotation. However, the contralateral external obliques and the ipsilateral erector spinae, and the ipsilateral latissimus dorsi came on before the actual angular rotation had begun. These muscles were therefore the major initiators and sustainers of the rotary motion. Contrary to an intuitive expectation the ipsilateral internal oblique had delayed onset. However, biomechanically, a coordinated simultaneous activity of contralateral external oblique and ipsilateral internal oblique could cause sufficient flexor moment to interfere with the axial rotation. The activation of contralateral external oblique and ipsilateral latissimus dorsi would create a rotary force couple maintaining the axis of rotation. The variable magnitude of the time of actual initiation of EMG activities in these muscles can be significantly affected by the initial position of the spine in transverse plane. Because of the varying amount of laxity and the magnitude of the neutral zone, the initial starting position can significantly vary the timing of the muscle firing. It is suggested that such has been the case. However, after traversing certain proportion of the neutral zone, the EMG activity will become considerably more perceptible. Closer to the middle of the neutral zone, even a slight activity of one or more muscles, whether primary initiator or not, can cause initial rotation, observing a definite pattern of muscle recruitment during the activity. The pattern recognition may be somewhat inaccurate in the delineation of the contribution of internal and external oblique. Despite careful placement of the electrodes on the internal obliques in the region of no overlap, a cross talk from external oblique from the same side cannot be entirely ruled out. In terms of the total contribution, the initiators and sustainer (the contralateral external oblique and the ipsilateral erector spinae and latissimus dorsi) contributed 65% of total muscle output in active rotation from 0 position to the end of range of axial rotation. The remaining 35% was offered by the other muscles that acted as antagonistic and stabilizing muscles. It was interesting to note that the same muscles were most active during the return phase accounting for approximately 50-52% of the total muscle output. Although the exact magnitude contribution is not entirely verifiable because of the limitation of technique, the trend is expected to be reliable. It would, therefore, appear that the recovery of the neutral position is entirely powered by the elastic recoil, which is controlled by the primary

22 Page 22 of 22 rotators. References 1. Andersson GBJ, Ortengren R, Nachemson AL, Schultz AB. Biomechanical analysis of loads on the lumbar spine in sitting and standing postures. Vol. VIII-A, Biomechanics. Champaign, Illinois: Human Kinetics Publishers, 1983: [Context Link] 2. Basmajian JV. Muscles alive: The functions revealed by electromyography. Baltimore: Williams and Wilkins, 1978: [Context Link] 3. Carlsoon S. The static muscle load in different work postures: An electromyographic study. Ergonomics 1961;4: [Context Link] 4. Kumar S. An electromyographic study of unresisted trunk rotation with normal velocity among normal males. Proceedings of IEEE. EMBS, San Diego, Oct 28-31, [Context Link] 5. Kumar S. Isolated planar trunk strength mobility measurement for the impaired and normal backs: Part I-The devices. Int J Indust Ergonomics 1996;17: [Context Link] 6. Kumar S, Garand D. Variation in lifting strength due to mechanical disadvantage and asymmetry of posture in isometric and isokinetic efforts. Proceedings of the 11th Trienniel International Ergonomics Association Conference, Paris 1991, Vol. 1, 84-6, [Context Link] 7. Kumar S, Panjabi MM. Neutral zone of thoraco-lumbar spine in axial rotation in vivo. Proceedings of International Society for the Study of Lumbar Spine; 216-7, [Context Link] 8. Kumar S, Panjabi MM. Axial rotation and spinal laxity (neutral zone) in vivo. J. Spinal Disorders 1995;8: [Context Link] 9. Marras W, Mirka M. A comprehensive evaluation of trunk response to asymmetric trunk motion. Spine 1992;17: Bibliographic Links [Context Link] 10. McGill S. Electromyographic activity of the abdominal and low back musculature during the generation of isometric and dynamic axial trunk torque: Implications for lumbar mechanics. J Orthop Res 1991;9: Bibliographic Links [Context Link] 11. Morris JM, Benner G, Lucas JB. An electromyographic study of the intrinsic muscles of the back in man. J Anat 1962;96: [Context Link] 12. Pope MH, Andersson GBJ, Broman H, Svensonn M, Zetterberg C. Electromyographic studies of the lumbar trunk musculature during the development of axial torques. J Ortho Res 1986;4: [Context Link] 13. Pope MH, Svensonn M, Andersson GBJ, Broman H, Zetterberg C. The role of prerotation of the trunk in axial twisting efforts. Spine 1987;12: Bibliographic Links [Context Link] 14. White AA, Panjabi MM. Clinical biomechanics of the spine. 2nd ed. Philadelphia: JB Lippincott, 1990:88. [Context Link] Key words: axial rotation; electromyographic of torso muscles in twisting; spinal rotation mechanism; trunk twisting Accession Number: Copyright (c) Ovid Technologies, Inc. Version: rel9.3.0, SourceID