- Email: [email protected]
Kinesiological research: The use of surface electromyography for assessing the effects of spinal manipulation
Journal of Electromyography and Kinesiology 22 (2012) 692–696
Contents lists available at SciVerse ScienceDirect
Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin
Review
Kinesiological research: The use of surface electromyography for assessing the effects of spinal manipulation Greg Lehman Private Practice, 1508 Queen St. E., Toronto, ON, Canada M4L 1E3
a r t i c l e
i n f o
Keywords: EMG Low back pain Spinal manipulation
a b s t r a c t Decreasing an elevated muscle tone is an often cited benefit of spinal manipulation. Spinal manipulation is theorized to disrupt an assumed pain-spasm-pain cycle that sufferers of low back pain may be experiencing. The current research has mostly investigated the short term influence of a single spinal manipulation on paraspinal muscle activity either at rest (e.g. standing or prone) or during simple movements (e.g. forward bend). The higher quality experiments to date have typically reported both reductions in muscle activity during lying prone or during the fully flexed position of forward bend. The only study measuring the long term influence of spinal manipulation has failed to document any change in muscle activity as measured with surface electromyography. Both manually delivered manipulations and manipulations delivered via a mechanical adjusting device have been associated with changes in muscle activation. Changes in muscle activity at muscles distant from the spinal joints manipulated (e.g. muscles in the upper limbs) have been documented following a single spinal manipulation however rather than the typical reduction in muscle activity an increase in resting activation has been reported. The state of muscle dysfunction (e.g. palpably tender or subjectively taut) may be a factor in achieving a myoelectric response to spinal manipulation. Currently, the clinical significance of short term changes in electromyographic amplitude following manipulation is unknown. Ó 2012 Elsevier Ltd. All rights reserved.
1. Background Surface electromyography (EMG) is proposed to give insight into the possible mechanisms of spinal manipulative therapy. Surface electromyography ‘‘is an experimental technique concerned with the development, recording and analysis of myoelectric signals. Myoelectric signals are formed by physiological variations in the state of muscle fiber membranes’’ (Basmajian and De Luca, 1985). The myoelectric signal recorded at the two surface electrodes is the summation of all of the motor unit action potentials within the area of the detecting electrodes. This interference pattern is subsequently differentially amplified, typically band passed filtered (e.g. removing high >1000 Hz and low <10 Hz signals) and converted from an analog to a digital signal during acquisition. Following this analog to digital conversion the raw myoelectric signal can undergo further processing depending on the needs of the researchers. In its most basic form measuring the activation amplitude of the raw or smoothed myoelectric signal is used to infer how electrically active a muscle is. Information on muscle activation amplitude levels has been used in an attempt to determine whether dysfunctional or altered motor control patterns occur in individuals with musculoskeletal
E-mail address: [email protected] 1050-6411/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.jelekin.2012.02.010
dysfunction. For example, a common finding in the erector spinae muscles during full forward flexion is myoelectric silence, where activation amplitude is significantly reduced. This myoelectric silence is termed as the flexion relaxation response (Neblett et al., 2003). In individuals with low back injury or pain this flexion relaxation response has been shown to be absent (Alschuler et al., 2009). Some researchers have suggested that the amplitude of the myoelectric signal may provide some insight into the pain-spasm-pain (i.e. vicious circle) theory of musculoskeletal dysfunction (van Dieën et al., 2003). Spinal manipulation research investigating the amplitude of the myoelectric signal appears to be driven by a clinical assumption that spinal manipulation can interrupt the processes of the pain-spasm-pain cycle in patients. Briefly, the pain-spasm-pain (vicious circle) model of musculoskeletal dysfunction contends that pain causes an increase in muscle activity which in turn causes further pain. This increased muscle activity induces ischemia from vascular compromise and becomes a source of further pain due to accumulation of painful metabolites (Travell et al., 1942). Johansson and Sojka (1991) described a detailed motor control pathway model for this vicious circle of musculoskeletal pain. The authors suggested that during muscle contraction metabolic products are produced (interstitial potassium, lactic acid and arachidonic acid) along with muscle ischemia that activate group III and IV muscle afferents. These muscle afferents subsequently excite the gamma motoneuron pool in the spinal cord which in
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
turn increases the sensitivity of the primary muscle spindle afferents. This increased muscle spindle sensitivity raises the activation level of the alpha motoneuron pool which leads to an increase in reflexively mediated muscle stiffness and potentially leads to an increase in the metabolic products/muscle ischemia thus completing the ‘‘vicious circle’’. While the validity of the pain-spasm-pain model has been questioned (van Dieën et al., 2003) it has served as a useful construct in an attempt to document what effects spinal manipulation has on muscular function. Theories on spinal manipulation by both Wyke (1985) and Korr (1975) suggest that spinal manipulation influences the excitability of the alpha-motoneuron pool. Briefly, they contend via various mechanisms that joint afferents are stimulated by the spinal manipulation and alpha-motoneuron excitability is decreased, thus a decrease in the resting tone or ‘‘spasm’’ of the affected muscle occurs. This theory of spinal manipulation influencing muscle activation amplitude has been investigated by numerous researchers. The following review will describe the research to date on the effects of spinal manipulation on surface myoelectric amplitude.
2. Research on static muscular function The majority of research has investigated the immediate effect of a single spinal manipulation or a single session of spinal manipulation on the activity of paraspinal muscles during static or simple dynamic tasks. Muscle activity is typically recorded while standing or recumbent immediately preceding and immediately following a spinal manipulation. One of the earliest published works was conducted by Shambaugh (1987), who measured the myoelectric signal from three pairs of paraspinal electrodes placed bilaterally over the upper trapezius, lower erector spinae and upper erector spinae immediately before and immediately after several spinal manipulations delivered to various sites of the spine. The authors did not control nor report which sites of the spine received spinal manipulation but the study reports that spinal manipulation was delivered at motion segments deemed hypomobile between the atlas and the pelvis. All spinal manipulations were delivered in the prone position. The myoelectric signal from the six recording sites was measured while the subjects were lying prone before the spinal manipulations and while lying prone two minutes after the last spinal manipulation was delivered. Control participations received motion palpation to assess the motion of all vertebral motion segments and then were asked to lie prone while muscle activity was recorded before and after a 5 min interval. No sham manipulation was delivered. Shambaugh reported a 25% average decrease in the myoelectric amplitude in 20 subjects with no reduction in the 14 individuals in the control group. Following an experimental setup similar to Shambaugh (1987), Lehman et al. (2001) investigated paraspinal muscle activity in low back pain sufferers (n = 17) while prone and during quiet stance immediately before and after a spinal manipulation. However, rather than studying the same muscle locations across subjects the muscle locations were determined by identifying motion segments that the subjects perceived as tender upon palpation and at non-tender motion segments distant (5 levels) from the tender motion segments. The study attempted to determine how spinal manipulation (directed at the painful motion segment) influenced the paraspinal muscle response to a painful mechanical pressure produced at tender motion segments and non-tender motion segments. While the muscle activity was collected in the prone position participants received a manually delivered posterior to anterior force (via an algometer at a force level perceived as painful) against the spinous process of the painful and then the non-
693
painful motion segment. The authors found that the application of a painful stimulus increased muscle activity in muscles adjacent to the painful motion segment and that this increase was decreased following spinal manipulation. This pattern of increasing muscle activity while a painful force was applied and a subsequent reduction in this myoelectric response following spinal manipulation was not seen at motion segments deemed non-painful. While a statistically significant decrease in muscle activity was seen, the qualitative assessment of the results suggests that reductions in muscle activity were not always present. The authors found that out of 36 muscle groups adjacent to a painful motion segment only 17 muscle groups exhibited decreases in muscle activity greater than 20%. This finding led the authors to suggest that a robust response to manipulation is variable across participants and may be patient or situation dependent. Further, this study is significantly limited by a lack of control group or control manipulation. In a related study, Lehman and McGill (2001) evaluated the influence of a single spinal manipulation on erector spinae (at T9 and L3 bilaterally) and abdominal (rectus abdominis and external oblique bilaterally) muscle activity during quiet stance in 12 subjects with persistent low back pain. The authors concluded that the majority of muscles studied did not demonstrate changes in the average amplitude of a rectified and smoothed (i.e. linear enveloped) EMG signal during quiet stance. However, incidents of change in muscle activity did occur. Across all muscles studied there were 17 incidents of changes in muscle activity with 16 of the changes being decreases. Of the 12 participants studied eight participants had at least one muscle show a change in muscle activity. The average change in muscle activity (of those muscles showing a change) was 24.39% reduction. Again, this study had no control group and results should be interpreted accordingly. In contrast to the variable response to spinal manipulation found in the previous studies, DeVocht et al. (2005) studied the immediate short term effects of spinal manipulation at paraspinal muscle sites that were deemed palpably taut in low back pain sufferers and found consistent decreases in the surface EMG signal amplitude. These authors investigated two types of spinal manipulation; (1) manually delivered high velocity low amplitude spine manipulation (n = 8) and (2) high velocity low amplitude manipulation delivered via a mechanical adjusting device (Activator Adjusting Instrument II, n = 8). Myoelectric signal measurements were taken both pre and post the different manipulations (within 5 min of the manipulation) while participants lay prone. A median percentage decrease of 39% occurred in the palpably taut muscle sites when using the Activator adjusting tool (11/15 muscles decreased 25% or more) while a median decrease of 43% was reported in muscle sites treated with high velocity low amplitude spine manipulation (13/16 muscle sites decreased 25% or more). The majority of research investigating surface EMG changes following spinal manipulation has measured activity during tasks where the myoelectric signal is typically not volitionally controlled. In other words, the muscle demand is minimal (e.g. at rest) or determined by the physical requirements of the task (e.g. bending forward). However, one experiment has investigated the influence of spine manipulation on the amplitude of surface electromyogram during a maximum voluntary static contraction. Keller and Colloca (2000) assessed the influence of a spine manipulation on erector spinae muscle activity (n = 40) during the performance of a maximal isometric back extension exertion in subjects with low back pain (n = 20). A second group of low back pain participants (n = 20) received either a sham manipulation or a control procedure. The spine manipulation in the experimental group was delivered with a mechanical adjusting device. Surface, linear-enveloped EMG was recorded from the erector spinae musculature at L3 and L5 during the trunk extension procedure. The authors found that 19 of the 20 patients in the spinal manipulation treatment group showed a
694
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
positive increase in the surface EMG amplitude during the MVC (average = 21%, range = 9.7% to 66.8%). There were no significant changes in the surface electromyographic amplitude in either the sham spine manipulation group or the control group. The mechanism underlying the increase is unknown. At least two possible explanations exist. First, spinal manipulation may have altered the excitability of the alpha-motoneuron pool resulting in increases in surface activity. Conversely, no alteration in alpha-motoneuron pool activity may have occurred. Rather, participants may have experienced less pain following the manipulation and this may have led to a reduction in a self imposed inhibition of effort resulting in greater muscle activity. 3. Research on dynamic muscular function During the full trunk flexion phase of forward bending a period of myoelectric silence or reduction occurs in the paraspinal muscle groups. This myoelectric silence has been termed the flexion relaxation phenomenon (FRP) and is often absent in individuals with low back pain. Using this simple forward bending task different measurement variables have been devised to describe the electrical activity that occurs during trunk forward flexion and the return to the upright standing. Researchers have evaluated the absolute level of the muscle activity that occurs during the different phases of movement (the forward bending phase, full forward flexion and re-extension to upright stance), the timing of muscle activity cessation and onset as a percent of the movement cycle or in terms of trunk flexion angle, and various ratios of muscle activity that can be calculated from the different phases of the activity. The different ratios that can be a calculated are; (1) flexion–relaxation ratio (FRR), calculated by dividing the muscle activity found during the flexion phase by the average muscle amplitude found during the full flexion phase; (2) the extension relaxation ratio (ERR) is calculated by dividing the maximal relative EMG during the extension phase by that found during full forward flexion; and (3) extension–flexion ratio (EFR) is calculated using the myoelectric activity during the extension and flexion phases. Lalanne et al. (2009) investigated the immediate influence of a spinal manipulation on the erector spinae activity recorded during forward flexion in experimental (n = 14) and control group (n = 13) of chronic low back pain patients. EMG responses in the experimental group were studied following a side posture, high velocity low amplitude spinal manipulation delivered to the middle lumbar spine. In the control group participants underwent a similar positioning setup but did not receive the high velocity low amplitude thrust. EMG data were recorded bilaterally from the erector spinae at the level of the L2 and L5 spinous processes. Data were averaged at each level. The authors assessed three variables; (1) the average trunk flexion angle found at the onset and cessation of the myoelectric silence of the flexion relaxation phase; (2) root mean squared averaged EMG amplitude during the full flexion phase of the movement; (3) the FRR. Lalanne et al. (2009) found no difference in the average angle of trunk flexion (onset or cessation) of the flexion relaxation phenomenon at either muscle site following spinal manipulation. In respect to the average muscle activity found during the full forward flexion phase there was no difference found in average muscle activity at L5, however, at L2, the average muscle activity was statistically less in the experimental group following spinal manipulation. With respect to the flexion relaxation ratio (FRR) a significant increase in the FRR was seen at both muscle sites in the experimental group but not in the control group. Changes in this ratio can occur from either a decrease in the muscle activity during the full forward flexion phase, an increase in muscle activity while flexing forward and/or a combination of both. Unfortunately, the authors did not provide data regarding the average muscle acti-
vation changes during the forward flexion part of the movement. At L2 this change may be attributed to the decrease in muscle activity found during the full forward flexion. At L5, there was not a statistically significant decrease in muscle activity during the full forward flexion phase suggesting that an increase in the FRR may be attributable to an increase in muscle activity recorded during the flexion phase of movement. However, the influence of the evident trend of decreased muscle activity during the full forward flexion phase cannot be discounted. This trend coupled with a possible slight increase in muscle activity during the forward flexion phase may account for the statistically significant finding of an increase in the flexion relaxation ratio. This finding of a statistically robust decrease (at least at one site in the spine) in myoelectric amplitude was not mirrored by the work of Lehman and McGill (2001) who found no significant change in 14 participants with low back pain following spinal manipulation at two spinal locations (T9 and L3). These authors did report variable responses across participants with some (5/ 14) participants showing a decrease in the myoelectric activity during full forward flexion. These authors also evaluated dynamic surface myoelectric activity from the erector spinae and abdominal muscles during planar movements (flexion, lateral bend and rotation). The authors found no systematic changes across subjects with low back pain rather they concluded that changes in EMG amplitude were variable with few incidents of change during dynamic simple planar movements. In contrast to the findings of Lehman and McGill (2001), Bicalho et al. (2010) investigated the influence of side posture, high velocity low amplitude spinal manipulation on the myoelectric signal amplitude of the paraspinal musculature at the level of L5 (n = 40) during trunk flexion and extension. The authors improved the experimental design of previous studies by adding a control group of low back pain patients. Ratios of the average activity across the different phases of the movement task were calculated pre and post manipulation. The ratios assessed were the flexion– relaxation ratio (FRR), the extension-relaxation ratio (ERR), and the extension–flexion ratio (EFR). The authors found that there was no change in muscle activity for either group during the dynamic flexion phase following manipulation – this finding being similar to all previous studies. However, a significant decrease in activity was found during the full forward flexion phase and during the extension phase in the experimental group following spinal manipulation. This change was not seen in the control group. A change in muscle activity had not previously been documented during flexion. With respect to the muscle activity ratios, a statistically significant increase was found for both the flexion relaxation ratio and the extension relaxation ratio in the experimental group but not the control. This finding is similar to the increase in the FRR found by Lalanne et al. (2009). Conversely, no change was found for the extension flexion ratio in either group. This lack of a change, coupled with the decrease in activity found during the full flexion phase suggests that the decrease in muscle activity during the full flexion phase may be more responsible for a change in the FRR and the EFR rather than any trend for muscle activity increase during the dynamic phase of trunk flexion–extension. One study to date has examined the long term (4 weeks) influence of repeated spinal manipulations on the paraspinal electromyographic signal during a forward bending task. Marshall and Murphy (2008) compared the extension relaxation ratio (although the ERR was called the flexion relaxation ratio in this study) before and after a 4 week course of either spinal manipulation therapy (n = 25) or non-spinal manipulation therapy (n = 25) in participants with low back pain. This component of the study was part of a larger study which incorporated an active rehabilitation program after the initial 4 weeks of passive therapy. The surface EMG signal was measured bilaterally at the levels of T12/L1 and L4/L5. After a 4 week trial
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
of spinal manipulation the authors found no differences in the ERR. Interestingly, the authors followed these participants for 1 year while the participants were involved in a spinal rehabilitation program. It was only during the incorporation of spinal exercises (performed either on a Swiss ball or on the ground) that changes in the ERR were noted, with changes being greatest in the group incorporating the Swiss ball into their rehabilitation.
4. Influence of spinal manipulation on distant muscle function The majority of studies have investigated the myoelectric signal from the paraspinal musculature with little work looking muscles other than the dorsal paraspinal muscles. Lehman and McGill (2001) reported on a single case of an internal oblique muscle demonstrating pain and which was assumed to be in ‘‘spasm’’ in one individual with low back pain. The experimenters measured the myoelectric signal of this muscle immediately pre–post a lumbar spinal manipulation and found an observable decrease (>30%) in resting muscle activity. However, this report was one simple case and was not studied in a systematic fashion. Addressing this lack in the literature Dunning and Rushton (2009) utilized a placebo-controlled, single blind repeated measures design to investigate the immediate influence of a right sided cervical spine (C5/6) high velocity low amplitude thrust manipulation on bilateral resting EMG amplitude in the biceps brachii musculature in asymptomatic participants (n = 54). Participants underwent a control procedure, a sham manipulation and a cervical spine manipulation. Myoelectric activity was recorded before and after these three conditions while participants lay prone on a treatment table. The sham manipulation consisted of moving the participants head and neck into a position just before a manipulative thrust is typically given however, no thrust was delivered and the head was then returned to neutral. For the control condition the participant received no manual contact. Data were collected for 30 s both before and after each condition. The order of conditions (six sequencing orders available) was randomly assigned to all participants. An 8 min ‘‘wash out’’ period occurred between all conditions in an attempt to minimize any carryover between the three conditions. Statistical analysis showed a significant difference between all three conditions for both left and right biceps brachii with increases in activity seen following spinal manipulation but not following the sham or control procedure. The mean percentage change of resting EMG activity of the right biceps brachii in the three conditions was 4.18% (control), 21.12% (sham), and 94.20% (high velocity low amplitude manipulation); and 2.16%, 17.15% and 80.04%, respectively, for the left biceps brachii. Ninety-four percentage of participants (51/54) showed increases in the biceps brachii with 6% (3/54) showing decreases in muscle activity. The mean percentage change in resting EMG activity following spinal manipulation to the right C5/6 segment was 94.20% and 80.04% for the right and left biceps brachii muscles, respectively. The right biceps brachii muscle experienced a statistically significant greater increase in resting muscle activity than the left. A secondary aim of the study evaluated the relevance of the audible ‘‘pop’’ or cavitation often associated with spinal manipulation. The authors compared the change in resting EMG activation levels in participants demonstrating an audible cavitation versus participants where no cavitation was heard. The authors reported that 32 of the 54 subjects demonstrated joint cavitation following the spine manipulation. The mean percentage change in resting EMG activity of the right biceps brachii muscle post spine manipulation was 79.79% and 115.16% for the cavitation and no cavitation groups, respectively. Similarly, the mean percentage change for the left biceps brachii muscle post manipulation was 69.61% and 95.20% for the cavitation and no cavitation groups,
695
respectively. Differences between groups were deemed statistically significant. 5. Repeatability of the myoelectric signal Factors other than neural drive can influence the amplitude of the EMG signal thus making comparisons across people, tasks, days and muscle groups often difficult (Lehman and McGill, 1999). Additionally, changes in the amplitude of the signal can be seen within an individual performing repeated tasks when slight differences in that individual’s posture or form occur. Repeatability of both the measurement tool (i.e. EMG) and the controls placed on the tasks evaluated are therefore important to consider whether a change in the EMG amplitude is a result of a specific intervention or due to inherent variability in the experiment. Watson et al. (1997) studied the repeatability of the amplitude of the signal during quiet standing, forward flexion, full forward flexion and the flexion–relaxation ratio from the upper and lower erector spinae muscle groups (n = 11) in a low back injured population. The authors concluded that the tasks were highly repeatable with reliability coefficients alpha ranging between .78 and .98 for within session repeatability. Across days repeatability saw a range of .58–.91. Owens et al. (2011) found similarly high within session repeatability values (ICC range = .86–.94) for the erector spinae’s flexion– relaxation response in a group of low back pain sufferers (n = 135). The reported relatively high repeatability measures occurred with strict controls during the movements studied. Many of the experiments reported to date do not describe the controls to ensure a consistent movement or posture studied before and after manipulation. Further, some movement variability may be unavoidable (e.g. forward bending at slightly different speeds across trials, variations in a participant’s posture may occur after lying prone on a treatment table) thus some caution should be exercised when inferring that changes can be attributed to spinal manipulation in the research studies lacking a control group, specifically the studies evaluating the myoelectric amplitude during movement. However, many of the experiments which found the most robust and statistically significant changes in myoelectric amplitude had control groups receiving an identical experimental set up with the spinal manipulation the only missing variable. No changes in the amplitude of the myoelectric signal were reported in these control groups. One limit of these control groups was the lack of sham manipulation so attributing the changes solely to a spinal manipulation should be done somewhat cautiously as it is possible that merely touching the patient or the patient knowing that they received a manipulation could influence the results. Regardless, the addition of control groups can adequately address the variability of the myoelectric signal and movement tasks. Further, many of the more robust and consistent changes in the myoelectric signal following spinal manipulation occurred during tasks that had no movement (e.g. lying prone or full forward flexion) suggesting that movement variability in these studies may not be a factor in the repeatability of the EMG signal. 6. Discussion on findings and limitations One difficulty in this form of research is the non-homogeneity of the low back pain populations studied. The muscles evaluated may not be dysfunctional and expecting changes following a spinal manipulation may be unwarranted. Changes in muscle activation appear to be most noticeable in muscles presenting with palpably tautness or pain. Interestingly, the assumed benefits or requirement of an audible ‘‘pop’’ or cavitation during spine manipulation does not appear necessary to have changes in the myoelectric amplitude. Further, the most common manipulation is a high velocity low amplitude thrust delivered manually – this form of delivery does
696
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
not appear to be a requirement to see changes in muscle activity as other forms of manipulation delivered via a mechanical force adjusting tool have also been associated with reductions in paraspinal activity as measured with surface EMG. The research to date has not demonstrated a possible mechanism or clinical significance of the changes documented. Considering the possibility that the elevated muscle activity during the full flexion phase of forward bending may be a positive adaptation to pain or soft tissue damage (e.g. ligamentous strain) it is plausible that short term decreases in muscle activity during full forward flexion are not beneficial. Preliminary long term research (i.e. 4 weeks) suggests that no change in muscle activation occurs following spinal manipulation yet clinical improvements are still evident during this time period suggesting a lack of relationship between clinical improvement and surface EMG measures of muscle function. While a disruption in the pain-spasm-pain cycle has been posited to occur following spinal manipulation this change may not be effectively robust considering that changes in amplitude are not seen when the myoelectric amplitude is the greatest (e.g. during movement). Rather, the greatest change in amplitude occurs during tasks when close to electrical silence is typically expected (e.g. full forward flexion, prone lying). Manipulation does not appear to change the muscle activation that is required to perform dynamic tasks of daily living that may subject an individual to high tissue stress or altered joint loading. There is no suggestion in the literature that an altered complex motor control program that might be associated with excessive (and possibly detrimental) muscle activation (e.g. trunk muscle co-contraction) can be positively influenced by spine manipulation. The research to date is relatively simple and future work may wish to address complex tasks or simple tasks while measuring complex activation profiles or spinal loading. Building on the work published currently, an interesting avenue of future work may wish to involve myoelectric measures in the formulation of clinical prediction rules. For example, is a robust response (e.g. large decrease in muscle amplitude or a substantial change in muscle onset timing) clinically correlated with long term improvements in disability and function. 7. Conclusions Spinal manipulation appears associated with short term changes in the amplitude of the myoelectric signal. The research studies with better control procedures and larger sample sizes consistently conclude that when changes in the myoelectric signal are present those changes are typically reductions in muscle activity when the muscles measured are the paraspinal muscles. However, a muscle having a close proximity to the manipulated segment, assumed to be transiently stretched by the manipulation, is not a necessary requirement to see changes in myoelectric amplitude following a spinal manipulation. Extremity muscles at a distance to the targeted spinal level of the manipulation have been shown to respond with changes in their myoelectric amplitude following spinal manipulation however, rather than a decrease in activity as is typically seen in axial muscles an increase in activation has been reported. Changes in muscle activity following spinal manipulation are most readily seen during simple, static tasks (e.g. lying prone) or during the non-movement portion of dynamic tasks (e.g. the fully flexed position of the lumbar spine during stance). In summary, the surface electromyogram has been used to document changes in muscle function following a spinal manipulation yet the significance and mechanism of these changes is mostly unknown. Other experimental designs and measures of muscle function may better afford a glimpse of the possible therapeutic or physiological mechanism of spinal manipulation. Conflict of interest None.
Acknowledgements No funding was received for the preparation of this manuscript. References Alschuler KN, Neblett R, Wiggert E, Haig AJ, Geisser ME. Flexion–relaxation and clinical features associated with chronic low back pain: a comparison of different methods of quantifying flexion–relaxation. Clin J Pain 2009;25(9):760–6. Basmajian JV, De Luca CJ. Muscles alive their function revealed by electromyography. Baltimore: Williams Wilkins; 1985. Bicalho E, Setti JA, Macagnan J, Cano JL, Manffra EF. Immediate effects of a highvelocity spine manipulation in paraspinal muscles activity of nonspecific chronic low-back pain subjects. Man Ther 2010;15(5):469–75. DeVocht JW, Pickar JG, Wilder DG. Spinal manipulation alters electromyographic activity of paraspinal muscles: a descriptive study. J Manipulative Physiol Ther 2005;28(7):465–71. Dunning J, Rushton A. The effects of cervical high-velocity low-amplitude thrust manipulation on resting electromyographic activity of the biceps brachii muscle. Man Ther 2009;14(5):508–13. Johansson H, Sojka P. Pathophysiological mechanisms involved in genesis and spread of muscular tension in occupational muscle pain and in chronic musculoskeletal pain syndromes: a hypothesis. Med Hypotheses 1991;35(3):196–203. Keller TS, Colloca CJ. Mechanical force spinal manipulation increases trunk muscle strength assessed by electromyography: a comparative clinical trial. J Manipulative Physiol Ther 2000;23(9):585–95. Korr IM. Proprioceptors and somatic dysfunction. J Am Osteopath Assoc 1975;74(7):638–50 [Review]. Lalanne K, Lafond D, Descarreaux M. Modulation of the flexion–relaxation response by spinal manipulative therapy: a control group study. J Manipulative Physiol Ther 2009;32(3):203–9. Lehman GJ, McGill SM. The importance of normalization in the interpretation of surface electromyography: a proof of principle. J Manipulative Physiol Ther 1999;22(7):444–6. Lehman GJ, McGill SM. Spinal manipulation causes variable spine kinematic and trunk muscle electromyographic responses. Clin Biomech (Bristol, Avon) 2001;16(4):293–9. Lehman GJ, Vernon H, McGill SM. Effects of a mechanical pain stimulus on erector spinae activity before and after a spinal manipulation in patients with back pain: a preliminary investigation. J Manipulative Physiol Ther 2001;24(6):402–6. Marshall PW, Murphy BA. Muscle activation changes after exercise rehabilitation for chronic low back pain. Arch Phys Med Rehabil 2008;89(7):1305–13. Neblett R, Mayer TG, Gatchel RJ, Keeley J, Proctor T, Anagnostis C. Quantifying the lumbar flexion–relaxation phenomenon: theory, normative data, and clinical applications. Spine (Phila Pa 1976) 2003;28(13):1435–46. Owens Jr EF, Gudavalli MR, Wilder DG. Paraspinal muscle function assessed with the flexion–relaxation ratio at baseline in a population of patients with back-related leg pain. J Manipulative Physiol Ther 2011;34(9):594–601 [Epub 2011 Jun 29]. Shambaugh P. Changes in electrical activity in muscles resulting from a chiropractic adjustment: a pilot study. J Manipulative Physiol Ther 1987;10(6):300–4. Travell J, Rinzter S, Herman M. Pain and disability of the shoulder and arm. JAMA 1942;120:417–22. van Dieën JH, Selen LP, Cholewicki J. Trunk muscle activation in low-back pain patients, an analysis of the literature. J Electromyogr Kinesiol 2003;13(4):333–51 [Review]. Watson PJ, Booker CK, Main CJ, Chen AC. Surface electromyography in the identification of chronic low back pain patients: the development of the flexion relaxation ratio. Clin Biomech (Bristol, Avon) 1997;12(3):165–71. Wyke BD. Articular neurology and manipulative therapy. In: Glasgow EF, editor. Aspects of manipulative therapy. New York: Churchill Livingstone; 1985.
Greg Lehman is currently in private practice as both a chiropractor and a physiotherapist in Toronto, Canada. Previously he was an assistant professor at the Canadian Memorial Chiropractic College and Research Scientist at the University of Waterloo where he received an M.Sc. in Spine Biomechanics (1999). Greg’s research has focused on spine and exercise biomechanics with special interest in muscular function as measured with surface electromyography. Greg maintains an active research blog at www.thebodymechanic.ca.
Contents lists available at SciVerse ScienceDirect
Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin
Review
Kinesiological research: The use of surface electromyography for assessing the effects of spinal manipulation Greg Lehman Private Practice, 1508 Queen St. E., Toronto, ON, Canada M4L 1E3
a r t i c l e
i n f o
Keywords: EMG Low back pain Spinal manipulation
a b s t r a c t Decreasing an elevated muscle tone is an often cited benefit of spinal manipulation. Spinal manipulation is theorized to disrupt an assumed pain-spasm-pain cycle that sufferers of low back pain may be experiencing. The current research has mostly investigated the short term influence of a single spinal manipulation on paraspinal muscle activity either at rest (e.g. standing or prone) or during simple movements (e.g. forward bend). The higher quality experiments to date have typically reported both reductions in muscle activity during lying prone or during the fully flexed position of forward bend. The only study measuring the long term influence of spinal manipulation has failed to document any change in muscle activity as measured with surface electromyography. Both manually delivered manipulations and manipulations delivered via a mechanical adjusting device have been associated with changes in muscle activation. Changes in muscle activity at muscles distant from the spinal joints manipulated (e.g. muscles in the upper limbs) have been documented following a single spinal manipulation however rather than the typical reduction in muscle activity an increase in resting activation has been reported. The state of muscle dysfunction (e.g. palpably tender or subjectively taut) may be a factor in achieving a myoelectric response to spinal manipulation. Currently, the clinical significance of short term changes in electromyographic amplitude following manipulation is unknown. Ó 2012 Elsevier Ltd. All rights reserved.
1. Background Surface electromyography (EMG) is proposed to give insight into the possible mechanisms of spinal manipulative therapy. Surface electromyography ‘‘is an experimental technique concerned with the development, recording and analysis of myoelectric signals. Myoelectric signals are formed by physiological variations in the state of muscle fiber membranes’’ (Basmajian and De Luca, 1985). The myoelectric signal recorded at the two surface electrodes is the summation of all of the motor unit action potentials within the area of the detecting electrodes. This interference pattern is subsequently differentially amplified, typically band passed filtered (e.g. removing high >1000 Hz and low <10 Hz signals) and converted from an analog to a digital signal during acquisition. Following this analog to digital conversion the raw myoelectric signal can undergo further processing depending on the needs of the researchers. In its most basic form measuring the activation amplitude of the raw or smoothed myoelectric signal is used to infer how electrically active a muscle is. Information on muscle activation amplitude levels has been used in an attempt to determine whether dysfunctional or altered motor control patterns occur in individuals with musculoskeletal
E-mail address: [email protected] 1050-6411/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.jelekin.2012.02.010
dysfunction. For example, a common finding in the erector spinae muscles during full forward flexion is myoelectric silence, where activation amplitude is significantly reduced. This myoelectric silence is termed as the flexion relaxation response (Neblett et al., 2003). In individuals with low back injury or pain this flexion relaxation response has been shown to be absent (Alschuler et al., 2009). Some researchers have suggested that the amplitude of the myoelectric signal may provide some insight into the pain-spasm-pain (i.e. vicious circle) theory of musculoskeletal dysfunction (van Dieën et al., 2003). Spinal manipulation research investigating the amplitude of the myoelectric signal appears to be driven by a clinical assumption that spinal manipulation can interrupt the processes of the pain-spasm-pain cycle in patients. Briefly, the pain-spasm-pain (vicious circle) model of musculoskeletal dysfunction contends that pain causes an increase in muscle activity which in turn causes further pain. This increased muscle activity induces ischemia from vascular compromise and becomes a source of further pain due to accumulation of painful metabolites (Travell et al., 1942). Johansson and Sojka (1991) described a detailed motor control pathway model for this vicious circle of musculoskeletal pain. The authors suggested that during muscle contraction metabolic products are produced (interstitial potassium, lactic acid and arachidonic acid) along with muscle ischemia that activate group III and IV muscle afferents. These muscle afferents subsequently excite the gamma motoneuron pool in the spinal cord which in
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
turn increases the sensitivity of the primary muscle spindle afferents. This increased muscle spindle sensitivity raises the activation level of the alpha motoneuron pool which leads to an increase in reflexively mediated muscle stiffness and potentially leads to an increase in the metabolic products/muscle ischemia thus completing the ‘‘vicious circle’’. While the validity of the pain-spasm-pain model has been questioned (van Dieën et al., 2003) it has served as a useful construct in an attempt to document what effects spinal manipulation has on muscular function. Theories on spinal manipulation by both Wyke (1985) and Korr (1975) suggest that spinal manipulation influences the excitability of the alpha-motoneuron pool. Briefly, they contend via various mechanisms that joint afferents are stimulated by the spinal manipulation and alpha-motoneuron excitability is decreased, thus a decrease in the resting tone or ‘‘spasm’’ of the affected muscle occurs. This theory of spinal manipulation influencing muscle activation amplitude has been investigated by numerous researchers. The following review will describe the research to date on the effects of spinal manipulation on surface myoelectric amplitude.
2. Research on static muscular function The majority of research has investigated the immediate effect of a single spinal manipulation or a single session of spinal manipulation on the activity of paraspinal muscles during static or simple dynamic tasks. Muscle activity is typically recorded while standing or recumbent immediately preceding and immediately following a spinal manipulation. One of the earliest published works was conducted by Shambaugh (1987), who measured the myoelectric signal from three pairs of paraspinal electrodes placed bilaterally over the upper trapezius, lower erector spinae and upper erector spinae immediately before and immediately after several spinal manipulations delivered to various sites of the spine. The authors did not control nor report which sites of the spine received spinal manipulation but the study reports that spinal manipulation was delivered at motion segments deemed hypomobile between the atlas and the pelvis. All spinal manipulations were delivered in the prone position. The myoelectric signal from the six recording sites was measured while the subjects were lying prone before the spinal manipulations and while lying prone two minutes after the last spinal manipulation was delivered. Control participations received motion palpation to assess the motion of all vertebral motion segments and then were asked to lie prone while muscle activity was recorded before and after a 5 min interval. No sham manipulation was delivered. Shambaugh reported a 25% average decrease in the myoelectric amplitude in 20 subjects with no reduction in the 14 individuals in the control group. Following an experimental setup similar to Shambaugh (1987), Lehman et al. (2001) investigated paraspinal muscle activity in low back pain sufferers (n = 17) while prone and during quiet stance immediately before and after a spinal manipulation. However, rather than studying the same muscle locations across subjects the muscle locations were determined by identifying motion segments that the subjects perceived as tender upon palpation and at non-tender motion segments distant (5 levels) from the tender motion segments. The study attempted to determine how spinal manipulation (directed at the painful motion segment) influenced the paraspinal muscle response to a painful mechanical pressure produced at tender motion segments and non-tender motion segments. While the muscle activity was collected in the prone position participants received a manually delivered posterior to anterior force (via an algometer at a force level perceived as painful) against the spinous process of the painful and then the non-
693
painful motion segment. The authors found that the application of a painful stimulus increased muscle activity in muscles adjacent to the painful motion segment and that this increase was decreased following spinal manipulation. This pattern of increasing muscle activity while a painful force was applied and a subsequent reduction in this myoelectric response following spinal manipulation was not seen at motion segments deemed non-painful. While a statistically significant decrease in muscle activity was seen, the qualitative assessment of the results suggests that reductions in muscle activity were not always present. The authors found that out of 36 muscle groups adjacent to a painful motion segment only 17 muscle groups exhibited decreases in muscle activity greater than 20%. This finding led the authors to suggest that a robust response to manipulation is variable across participants and may be patient or situation dependent. Further, this study is significantly limited by a lack of control group or control manipulation. In a related study, Lehman and McGill (2001) evaluated the influence of a single spinal manipulation on erector spinae (at T9 and L3 bilaterally) and abdominal (rectus abdominis and external oblique bilaterally) muscle activity during quiet stance in 12 subjects with persistent low back pain. The authors concluded that the majority of muscles studied did not demonstrate changes in the average amplitude of a rectified and smoothed (i.e. linear enveloped) EMG signal during quiet stance. However, incidents of change in muscle activity did occur. Across all muscles studied there were 17 incidents of changes in muscle activity with 16 of the changes being decreases. Of the 12 participants studied eight participants had at least one muscle show a change in muscle activity. The average change in muscle activity (of those muscles showing a change) was 24.39% reduction. Again, this study had no control group and results should be interpreted accordingly. In contrast to the variable response to spinal manipulation found in the previous studies, DeVocht et al. (2005) studied the immediate short term effects of spinal manipulation at paraspinal muscle sites that were deemed palpably taut in low back pain sufferers and found consistent decreases in the surface EMG signal amplitude. These authors investigated two types of spinal manipulation; (1) manually delivered high velocity low amplitude spine manipulation (n = 8) and (2) high velocity low amplitude manipulation delivered via a mechanical adjusting device (Activator Adjusting Instrument II, n = 8). Myoelectric signal measurements were taken both pre and post the different manipulations (within 5 min of the manipulation) while participants lay prone. A median percentage decrease of 39% occurred in the palpably taut muscle sites when using the Activator adjusting tool (11/15 muscles decreased 25% or more) while a median decrease of 43% was reported in muscle sites treated with high velocity low amplitude spine manipulation (13/16 muscle sites decreased 25% or more). The majority of research investigating surface EMG changes following spinal manipulation has measured activity during tasks where the myoelectric signal is typically not volitionally controlled. In other words, the muscle demand is minimal (e.g. at rest) or determined by the physical requirements of the task (e.g. bending forward). However, one experiment has investigated the influence of spine manipulation on the amplitude of surface electromyogram during a maximum voluntary static contraction. Keller and Colloca (2000) assessed the influence of a spine manipulation on erector spinae muscle activity (n = 40) during the performance of a maximal isometric back extension exertion in subjects with low back pain (n = 20). A second group of low back pain participants (n = 20) received either a sham manipulation or a control procedure. The spine manipulation in the experimental group was delivered with a mechanical adjusting device. Surface, linear-enveloped EMG was recorded from the erector spinae musculature at L3 and L5 during the trunk extension procedure. The authors found that 19 of the 20 patients in the spinal manipulation treatment group showed a
694
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
positive increase in the surface EMG amplitude during the MVC (average = 21%, range = 9.7% to 66.8%). There were no significant changes in the surface electromyographic amplitude in either the sham spine manipulation group or the control group. The mechanism underlying the increase is unknown. At least two possible explanations exist. First, spinal manipulation may have altered the excitability of the alpha-motoneuron pool resulting in increases in surface activity. Conversely, no alteration in alpha-motoneuron pool activity may have occurred. Rather, participants may have experienced less pain following the manipulation and this may have led to a reduction in a self imposed inhibition of effort resulting in greater muscle activity. 3. Research on dynamic muscular function During the full trunk flexion phase of forward bending a period of myoelectric silence or reduction occurs in the paraspinal muscle groups. This myoelectric silence has been termed the flexion relaxation phenomenon (FRP) and is often absent in individuals with low back pain. Using this simple forward bending task different measurement variables have been devised to describe the electrical activity that occurs during trunk forward flexion and the return to the upright standing. Researchers have evaluated the absolute level of the muscle activity that occurs during the different phases of movement (the forward bending phase, full forward flexion and re-extension to upright stance), the timing of muscle activity cessation and onset as a percent of the movement cycle or in terms of trunk flexion angle, and various ratios of muscle activity that can be calculated from the different phases of the activity. The different ratios that can be a calculated are; (1) flexion–relaxation ratio (FRR), calculated by dividing the muscle activity found during the flexion phase by the average muscle amplitude found during the full flexion phase; (2) the extension relaxation ratio (ERR) is calculated by dividing the maximal relative EMG during the extension phase by that found during full forward flexion; and (3) extension–flexion ratio (EFR) is calculated using the myoelectric activity during the extension and flexion phases. Lalanne et al. (2009) investigated the immediate influence of a spinal manipulation on the erector spinae activity recorded during forward flexion in experimental (n = 14) and control group (n = 13) of chronic low back pain patients. EMG responses in the experimental group were studied following a side posture, high velocity low amplitude spinal manipulation delivered to the middle lumbar spine. In the control group participants underwent a similar positioning setup but did not receive the high velocity low amplitude thrust. EMG data were recorded bilaterally from the erector spinae at the level of the L2 and L5 spinous processes. Data were averaged at each level. The authors assessed three variables; (1) the average trunk flexion angle found at the onset and cessation of the myoelectric silence of the flexion relaxation phase; (2) root mean squared averaged EMG amplitude during the full flexion phase of the movement; (3) the FRR. Lalanne et al. (2009) found no difference in the average angle of trunk flexion (onset or cessation) of the flexion relaxation phenomenon at either muscle site following spinal manipulation. In respect to the average muscle activity found during the full forward flexion phase there was no difference found in average muscle activity at L5, however, at L2, the average muscle activity was statistically less in the experimental group following spinal manipulation. With respect to the flexion relaxation ratio (FRR) a significant increase in the FRR was seen at both muscle sites in the experimental group but not in the control group. Changes in this ratio can occur from either a decrease in the muscle activity during the full forward flexion phase, an increase in muscle activity while flexing forward and/or a combination of both. Unfortunately, the authors did not provide data regarding the average muscle acti-
vation changes during the forward flexion part of the movement. At L2 this change may be attributed to the decrease in muscle activity found during the full forward flexion. At L5, there was not a statistically significant decrease in muscle activity during the full forward flexion phase suggesting that an increase in the FRR may be attributable to an increase in muscle activity recorded during the flexion phase of movement. However, the influence of the evident trend of decreased muscle activity during the full forward flexion phase cannot be discounted. This trend coupled with a possible slight increase in muscle activity during the forward flexion phase may account for the statistically significant finding of an increase in the flexion relaxation ratio. This finding of a statistically robust decrease (at least at one site in the spine) in myoelectric amplitude was not mirrored by the work of Lehman and McGill (2001) who found no significant change in 14 participants with low back pain following spinal manipulation at two spinal locations (T9 and L3). These authors did report variable responses across participants with some (5/ 14) participants showing a decrease in the myoelectric activity during full forward flexion. These authors also evaluated dynamic surface myoelectric activity from the erector spinae and abdominal muscles during planar movements (flexion, lateral bend and rotation). The authors found no systematic changes across subjects with low back pain rather they concluded that changes in EMG amplitude were variable with few incidents of change during dynamic simple planar movements. In contrast to the findings of Lehman and McGill (2001), Bicalho et al. (2010) investigated the influence of side posture, high velocity low amplitude spinal manipulation on the myoelectric signal amplitude of the paraspinal musculature at the level of L5 (n = 40) during trunk flexion and extension. The authors improved the experimental design of previous studies by adding a control group of low back pain patients. Ratios of the average activity across the different phases of the movement task were calculated pre and post manipulation. The ratios assessed were the flexion– relaxation ratio (FRR), the extension-relaxation ratio (ERR), and the extension–flexion ratio (EFR). The authors found that there was no change in muscle activity for either group during the dynamic flexion phase following manipulation – this finding being similar to all previous studies. However, a significant decrease in activity was found during the full forward flexion phase and during the extension phase in the experimental group following spinal manipulation. This change was not seen in the control group. A change in muscle activity had not previously been documented during flexion. With respect to the muscle activity ratios, a statistically significant increase was found for both the flexion relaxation ratio and the extension relaxation ratio in the experimental group but not the control. This finding is similar to the increase in the FRR found by Lalanne et al. (2009). Conversely, no change was found for the extension flexion ratio in either group. This lack of a change, coupled with the decrease in activity found during the full flexion phase suggests that the decrease in muscle activity during the full flexion phase may be more responsible for a change in the FRR and the EFR rather than any trend for muscle activity increase during the dynamic phase of trunk flexion–extension. One study to date has examined the long term (4 weeks) influence of repeated spinal manipulations on the paraspinal electromyographic signal during a forward bending task. Marshall and Murphy (2008) compared the extension relaxation ratio (although the ERR was called the flexion relaxation ratio in this study) before and after a 4 week course of either spinal manipulation therapy (n = 25) or non-spinal manipulation therapy (n = 25) in participants with low back pain. This component of the study was part of a larger study which incorporated an active rehabilitation program after the initial 4 weeks of passive therapy. The surface EMG signal was measured bilaterally at the levels of T12/L1 and L4/L5. After a 4 week trial
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
of spinal manipulation the authors found no differences in the ERR. Interestingly, the authors followed these participants for 1 year while the participants were involved in a spinal rehabilitation program. It was only during the incorporation of spinal exercises (performed either on a Swiss ball or on the ground) that changes in the ERR were noted, with changes being greatest in the group incorporating the Swiss ball into their rehabilitation.
4. Influence of spinal manipulation on distant muscle function The majority of studies have investigated the myoelectric signal from the paraspinal musculature with little work looking muscles other than the dorsal paraspinal muscles. Lehman and McGill (2001) reported on a single case of an internal oblique muscle demonstrating pain and which was assumed to be in ‘‘spasm’’ in one individual with low back pain. The experimenters measured the myoelectric signal of this muscle immediately pre–post a lumbar spinal manipulation and found an observable decrease (>30%) in resting muscle activity. However, this report was one simple case and was not studied in a systematic fashion. Addressing this lack in the literature Dunning and Rushton (2009) utilized a placebo-controlled, single blind repeated measures design to investigate the immediate influence of a right sided cervical spine (C5/6) high velocity low amplitude thrust manipulation on bilateral resting EMG amplitude in the biceps brachii musculature in asymptomatic participants (n = 54). Participants underwent a control procedure, a sham manipulation and a cervical spine manipulation. Myoelectric activity was recorded before and after these three conditions while participants lay prone on a treatment table. The sham manipulation consisted of moving the participants head and neck into a position just before a manipulative thrust is typically given however, no thrust was delivered and the head was then returned to neutral. For the control condition the participant received no manual contact. Data were collected for 30 s both before and after each condition. The order of conditions (six sequencing orders available) was randomly assigned to all participants. An 8 min ‘‘wash out’’ period occurred between all conditions in an attempt to minimize any carryover between the three conditions. Statistical analysis showed a significant difference between all three conditions for both left and right biceps brachii with increases in activity seen following spinal manipulation but not following the sham or control procedure. The mean percentage change of resting EMG activity of the right biceps brachii in the three conditions was 4.18% (control), 21.12% (sham), and 94.20% (high velocity low amplitude manipulation); and 2.16%, 17.15% and 80.04%, respectively, for the left biceps brachii. Ninety-four percentage of participants (51/54) showed increases in the biceps brachii with 6% (3/54) showing decreases in muscle activity. The mean percentage change in resting EMG activity following spinal manipulation to the right C5/6 segment was 94.20% and 80.04% for the right and left biceps brachii muscles, respectively. The right biceps brachii muscle experienced a statistically significant greater increase in resting muscle activity than the left. A secondary aim of the study evaluated the relevance of the audible ‘‘pop’’ or cavitation often associated with spinal manipulation. The authors compared the change in resting EMG activation levels in participants demonstrating an audible cavitation versus participants where no cavitation was heard. The authors reported that 32 of the 54 subjects demonstrated joint cavitation following the spine manipulation. The mean percentage change in resting EMG activity of the right biceps brachii muscle post spine manipulation was 79.79% and 115.16% for the cavitation and no cavitation groups, respectively. Similarly, the mean percentage change for the left biceps brachii muscle post manipulation was 69.61% and 95.20% for the cavitation and no cavitation groups,
695
respectively. Differences between groups were deemed statistically significant. 5. Repeatability of the myoelectric signal Factors other than neural drive can influence the amplitude of the EMG signal thus making comparisons across people, tasks, days and muscle groups often difficult (Lehman and McGill, 1999). Additionally, changes in the amplitude of the signal can be seen within an individual performing repeated tasks when slight differences in that individual’s posture or form occur. Repeatability of both the measurement tool (i.e. EMG) and the controls placed on the tasks evaluated are therefore important to consider whether a change in the EMG amplitude is a result of a specific intervention or due to inherent variability in the experiment. Watson et al. (1997) studied the repeatability of the amplitude of the signal during quiet standing, forward flexion, full forward flexion and the flexion–relaxation ratio from the upper and lower erector spinae muscle groups (n = 11) in a low back injured population. The authors concluded that the tasks were highly repeatable with reliability coefficients alpha ranging between .78 and .98 for within session repeatability. Across days repeatability saw a range of .58–.91. Owens et al. (2011) found similarly high within session repeatability values (ICC range = .86–.94) for the erector spinae’s flexion– relaxation response in a group of low back pain sufferers (n = 135). The reported relatively high repeatability measures occurred with strict controls during the movements studied. Many of the experiments reported to date do not describe the controls to ensure a consistent movement or posture studied before and after manipulation. Further, some movement variability may be unavoidable (e.g. forward bending at slightly different speeds across trials, variations in a participant’s posture may occur after lying prone on a treatment table) thus some caution should be exercised when inferring that changes can be attributed to spinal manipulation in the research studies lacking a control group, specifically the studies evaluating the myoelectric amplitude during movement. However, many of the experiments which found the most robust and statistically significant changes in myoelectric amplitude had control groups receiving an identical experimental set up with the spinal manipulation the only missing variable. No changes in the amplitude of the myoelectric signal were reported in these control groups. One limit of these control groups was the lack of sham manipulation so attributing the changes solely to a spinal manipulation should be done somewhat cautiously as it is possible that merely touching the patient or the patient knowing that they received a manipulation could influence the results. Regardless, the addition of control groups can adequately address the variability of the myoelectric signal and movement tasks. Further, many of the more robust and consistent changes in the myoelectric signal following spinal manipulation occurred during tasks that had no movement (e.g. lying prone or full forward flexion) suggesting that movement variability in these studies may not be a factor in the repeatability of the EMG signal. 6. Discussion on findings and limitations One difficulty in this form of research is the non-homogeneity of the low back pain populations studied. The muscles evaluated may not be dysfunctional and expecting changes following a spinal manipulation may be unwarranted. Changes in muscle activation appear to be most noticeable in muscles presenting with palpably tautness or pain. Interestingly, the assumed benefits or requirement of an audible ‘‘pop’’ or cavitation during spine manipulation does not appear necessary to have changes in the myoelectric amplitude. Further, the most common manipulation is a high velocity low amplitude thrust delivered manually – this form of delivery does
696
G. Lehman / Journal of Electromyography and Kinesiology 22 (2012) 692–696
not appear to be a requirement to see changes in muscle activity as other forms of manipulation delivered via a mechanical force adjusting tool have also been associated with reductions in paraspinal activity as measured with surface EMG. The research to date has not demonstrated a possible mechanism or clinical significance of the changes documented. Considering the possibility that the elevated muscle activity during the full flexion phase of forward bending may be a positive adaptation to pain or soft tissue damage (e.g. ligamentous strain) it is plausible that short term decreases in muscle activity during full forward flexion are not beneficial. Preliminary long term research (i.e. 4 weeks) suggests that no change in muscle activation occurs following spinal manipulation yet clinical improvements are still evident during this time period suggesting a lack of relationship between clinical improvement and surface EMG measures of muscle function. While a disruption in the pain-spasm-pain cycle has been posited to occur following spinal manipulation this change may not be effectively robust considering that changes in amplitude are not seen when the myoelectric amplitude is the greatest (e.g. during movement). Rather, the greatest change in amplitude occurs during tasks when close to electrical silence is typically expected (e.g. full forward flexion, prone lying). Manipulation does not appear to change the muscle activation that is required to perform dynamic tasks of daily living that may subject an individual to high tissue stress or altered joint loading. There is no suggestion in the literature that an altered complex motor control program that might be associated with excessive (and possibly detrimental) muscle activation (e.g. trunk muscle co-contraction) can be positively influenced by spine manipulation. The research to date is relatively simple and future work may wish to address complex tasks or simple tasks while measuring complex activation profiles or spinal loading. Building on the work published currently, an interesting avenue of future work may wish to involve myoelectric measures in the formulation of clinical prediction rules. For example, is a robust response (e.g. large decrease in muscle amplitude or a substantial change in muscle onset timing) clinically correlated with long term improvements in disability and function. 7. Conclusions Spinal manipulation appears associated with short term changes in the amplitude of the myoelectric signal. The research studies with better control procedures and larger sample sizes consistently conclude that when changes in the myoelectric signal are present those changes are typically reductions in muscle activity when the muscles measured are the paraspinal muscles. However, a muscle having a close proximity to the manipulated segment, assumed to be transiently stretched by the manipulation, is not a necessary requirement to see changes in myoelectric amplitude following a spinal manipulation. Extremity muscles at a distance to the targeted spinal level of the manipulation have been shown to respond with changes in their myoelectric amplitude following spinal manipulation however, rather than a decrease in activity as is typically seen in axial muscles an increase in activation has been reported. Changes in muscle activity following spinal manipulation are most readily seen during simple, static tasks (e.g. lying prone) or during the non-movement portion of dynamic tasks (e.g. the fully flexed position of the lumbar spine during stance). In summary, the surface electromyogram has been used to document changes in muscle function following a spinal manipulation yet the significance and mechanism of these changes is mostly unknown. Other experimental designs and measures of muscle function may better afford a glimpse of the possible therapeutic or physiological mechanism of spinal manipulation. Conflict of interest None.
Acknowledgements No funding was received for the preparation of this manuscript. References Alschuler KN, Neblett R, Wiggert E, Haig AJ, Geisser ME. Flexion–relaxation and clinical features associated with chronic low back pain: a comparison of different methods of quantifying flexion–relaxation. Clin J Pain 2009;25(9):760–6. Basmajian JV, De Luca CJ. Muscles alive their function revealed by electromyography. Baltimore: Williams Wilkins; 1985. Bicalho E, Setti JA, Macagnan J, Cano JL, Manffra EF. Immediate effects of a highvelocity spine manipulation in paraspinal muscles activity of nonspecific chronic low-back pain subjects. Man Ther 2010;15(5):469–75. DeVocht JW, Pickar JG, Wilder DG. Spinal manipulation alters electromyographic activity of paraspinal muscles: a descriptive study. J Manipulative Physiol Ther 2005;28(7):465–71. Dunning J, Rushton A. The effects of cervical high-velocity low-amplitude thrust manipulation on resting electromyographic activity of the biceps brachii muscle. Man Ther 2009;14(5):508–13. Johansson H, Sojka P. Pathophysiological mechanisms involved in genesis and spread of muscular tension in occupational muscle pain and in chronic musculoskeletal pain syndromes: a hypothesis. Med Hypotheses 1991;35(3):196–203. Keller TS, Colloca CJ. Mechanical force spinal manipulation increases trunk muscle strength assessed by electromyography: a comparative clinical trial. J Manipulative Physiol Ther 2000;23(9):585–95. Korr IM. Proprioceptors and somatic dysfunction. J Am Osteopath Assoc 1975;74(7):638–50 [Review]. Lalanne K, Lafond D, Descarreaux M. Modulation of the flexion–relaxation response by spinal manipulative therapy: a control group study. J Manipulative Physiol Ther 2009;32(3):203–9. Lehman GJ, McGill SM. The importance of normalization in the interpretation of surface electromyography: a proof of principle. J Manipulative Physiol Ther 1999;22(7):444–6. Lehman GJ, McGill SM. Spinal manipulation causes variable spine kinematic and trunk muscle electromyographic responses. Clin Biomech (Bristol, Avon) 2001;16(4):293–9. Lehman GJ, Vernon H, McGill SM. Effects of a mechanical pain stimulus on erector spinae activity before and after a spinal manipulation in patients with back pain: a preliminary investigation. J Manipulative Physiol Ther 2001;24(6):402–6. Marshall PW, Murphy BA. Muscle activation changes after exercise rehabilitation for chronic low back pain. Arch Phys Med Rehabil 2008;89(7):1305–13. Neblett R, Mayer TG, Gatchel RJ, Keeley J, Proctor T, Anagnostis C. Quantifying the lumbar flexion–relaxation phenomenon: theory, normative data, and clinical applications. Spine (Phila Pa 1976) 2003;28(13):1435–46. Owens Jr EF, Gudavalli MR, Wilder DG. Paraspinal muscle function assessed with the flexion–relaxation ratio at baseline in a population of patients with back-related leg pain. J Manipulative Physiol Ther 2011;34(9):594–601 [Epub 2011 Jun 29]. Shambaugh P. Changes in electrical activity in muscles resulting from a chiropractic adjustment: a pilot study. J Manipulative Physiol Ther 1987;10(6):300–4. Travell J, Rinzter S, Herman M. Pain and disability of the shoulder and arm. JAMA 1942;120:417–22. van Dieën JH, Selen LP, Cholewicki J. Trunk muscle activation in low-back pain patients, an analysis of the literature. J Electromyogr Kinesiol 2003;13(4):333–51 [Review]. Watson PJ, Booker CK, Main CJ, Chen AC. Surface electromyography in the identification of chronic low back pain patients: the development of the flexion relaxation ratio. Clin Biomech (Bristol, Avon) 1997;12(3):165–71. Wyke BD. Articular neurology and manipulative therapy. In: Glasgow EF, editor. Aspects of manipulative therapy. New York: Churchill Livingstone; 1985.
Greg Lehman is currently in private practice as both a chiropractor and a physiotherapist in Toronto, Canada. Previously he was an assistant professor at the Canadian Memorial Chiropractic College and Research Scientist at the University of Waterloo where he received an M.Sc. in Spine Biomechanics (1999). Greg’s research has focused on spine and exercise biomechanics with special interest in muscular function as measured with surface electromyography. Greg maintains an active research blog at www.thebodymechanic.ca.