Mechanical Force Spinal Manipulation Increases Trunk Muscle Strength Assessed By Electromyography: A Comparative Clinical Trial


Mechanical Force Spinal Manipulation Increases Trunk Muscle Strength Assessed By Electromyography: A Comparative Clinical Trial

Tony S. Keller, PhD, Christopher J. Colloca, DC

Vermont Orthopedic Biomechanics Consultants,
Burlington, Vt., USA.

OBJECTIVE:   The objective of this study was to determine whether mechanical force, manually-assisted (MFMA) spinal manipulative therapy (SMT) affects paraspinal muscle strength as assessed through use of surface electromyography (sEMG).

DESIGN:   Prospective clinical trial comparing sEMG output in 1 active treatment group and 2 control groups.

SETTING:   Outpatient chiropractic clinic, Phoenix, AZ.

SUBJECTS:   Forty subjects with low back pain (LBP) participated in the study. Twenty patients with LBP (9 females and 11 males with a mean age of 35 years and 51 years, respectively) and 20 age- and sex-matched sham-SMT/control LBP subjects (10 females and 10 males with a mean age of 40 years and 52 years, respectively) were assessed.

METHODS:   Twenty consecutive patients with LBP (SMT treatment group) performed maximum voluntary contraction (MVC) isometric trunk extensions while lying prone on a treatment table. Surface, linear-enveloped sEMG was recorded from the erector spinae musculature at L3 and L5 during a trunk extension procedure. Patients were then assessed through use of the Activator Methods Chiropractic Technique protocol, during which time they were treated through use of MFMA SMT. The MFMA SMT treatment was followed by a dynamic stiffness and algometry assessment, after which a second or post-MVC isometric trunk extension and sEMG assessment were performed. Another 20 consecutive subjects with LBP were assigned to one of two other groups, a sham-SMT group and a control group. The sham-SMT group underwent the same experimental protocol with the exception that the subjects received a sham-MFMA SMT and dynamic stiffness assessment. The control group subjects received no SMT treatment, stiffness assessment, or algometry assessment intervention. Within-group analysis of MVC sEMG output (pre-SMT vs post-SMT sEMG output) and across-group analysis of MVC sEMG output ratio (post-SMT sEMG/pre-SMT sEMG output) during MVC was performed through use of a paired observations t test (POTT) and a robust analysis of variance (RANOVA), respectively.

MAIN OUTCOME MEASURES:   Surface, linear-enveloped EMG recordings during isometric MVC trunk extension were used as the primary outcome measure.

RESULTS:   Nineteen of the 20 patients in the SMT treatment group showed a positive increase in sEMG output during MVC (range, -9.7% to 66.8%) after the active MFMA SMT treatment and stiffness assessment. The SMT treatment group showed a significant (POTT, P < 0.001) increase in erector spinae muscle sEMG output (21% increase in comparison with pre-SMT levels) during MVC isometric trunk extension trials. There were no significant changes in pre-SMT vs post-SMT MVC sEMG output for the sham-SMT (5.8% increase) and control (3.9% increase) groups. Moreover, the sEMG output ratio of the SMT treatment group was significantly greater (robust analysis of variance, P = 0.05) than either that of the sham-SMT group or that of the control group.

CONCLUSIONS:   The results of this preliminary clinical trial demonstrated that MFMA SMT results in a significant increase in sEMG erector spinae isometric MVC muscle output. These findings indicate that altered muscle function may be a potential short-term therapeutic effect of MFMA SMT, and they form a basis for a randomized, controlled clinical trial to further investigate acute and long-term changes in low back function.

From the Full-Text Article:


A number of different outcome measures have been used to investigate the effectiveness of spinal manipulative therapy (SMT). Among these are subjective measures of patient’s self-reported level of pain, disability, and functional status. [1-4] Objective measures used to assess outcome in patients receiving treatment for low back pain (LBP) include orthopedic examination tests,5 range of motion assessment, [6, 7] radiographic measures, [8, 9] spinal stiffness, [10] and trunk muscle strength. [11]

Increasing support for the role of muscles in stabilizing the spine have drawn attention in recent years. [12-17] The erector spinae musculature acts posterior to the intervertebral joint centers to compensate for the net moment caused by external load and body weight.14 The maintenance of posture and performance of purposeful trunk motion are the results of coordinated load-sharing between the passive discoligamentous tissues and active contractile muscular tissues to balance external loads. A disorder in the musculoskeletal system or loss of control in the neuromuscular system may result in excessive load-sharing of the passive system that can cause abnormal motion and increased deformation of the highly innervated structures of the spine.18 Abnormal motion and increased strain are often accompanied by increased pain and discomfort. [14, 19-22] Individuals with LBP have been found to use a different motor control strategy in comparison with asymptomatic healthy individuals, and this may be a result of pain or damage to muscular, ligamentous, or nervous (mechanosensitive) tissues. [23-27] The role of rehabilitation programs in improving objective outcomes, including increases in trunk muscle strength, mobility, and functional capacity as well as improved motor control system function, are important goals of patient care. [28]

Measurement of muscle action and forces may be documented by electromyographic studies. [29] Because muscle output has been found to be closely correlated with muscle strength, [30-32] the potential ability of rehabilitation programs and the role of SMT in affecting the neuromuscular system are of interest to researchers and clinicians. [33-35] To date, few studies have been done to investigate the effect of SMT on trunk muscle output.

In a recent prospective clinical study, we investigated the mechanical and muscular behavior of human lumbar spinal segments to high-loading-rate posterior-to-anterior (PA) manipulative thrusts in vivo. [36] As part of our experimental protocol, prone-lying patients were asked to perform isometric maximum voluntary contraction (MVC) trunk extension efforts while surface, linear-enveloped electromyographic electrodes monitored erector spinae muscular activity from leads placed bilaterally at L3 and L5. These maximums being used to normalize surface electromyography (sEMG) data, [37] reflex responses occurring during mechanical force, manually-assisted (MFMA) SMTs were calculated to establish relative muscle activity, expressed as a percentage of the maximum muscle activity.38 An unexpected outcome of this work was the finding that there was a consistent and significant increase in the post-SMT isometric MVC sEMG output in comparison with the pre-SMT output. This finding prompted us to undertake a clinical trial on a second group of 20 age- and sex-matched subjects. Our null hypothesis was that there would be no significant change in isometric MVC sEMG output for subjects undergoing sham-SMT treatment or a control protocol.


SMT is a commonly used conservative treatment in the care of patients with LBP. [4] Our aim was to determine whether MFMA SMT affects paraspinal muscle strength as assessed through use of sEMG. The role of muscle strength as an etiologic or predisposing factor of low back disorders is controversial. [28] Some studies [45, 46] have found little association between muscle strength and the development of low back symptoms, whereas others [47-49] have found that the trunk muscles in patients with chronic LBP are significantly weaker than those of healthy individuals. A muscle deficiency model of chronic LBP was proposed by Cassisi et al. [50] Those investigators found that patients with chronic LBP exhibited lower peak torque and lower maximum surface-integrated electromyography bilaterally during isometric extension. Triano and Schultz [11] found that both myoelectric signal levels and trunk strength ratios of patients with LBP were inversely related to the patients’ disability scores. Because paraspinal muscle function is thought to be essential for providing intrinsic support to the spinal column, sub-optimal or impaired muscle control and function could permit abnormal motion segment movement or predispose a person to pain. [51]

Electromyographic signal analysis, a commonly used technique for quantifying muscle function during lifting tasks, has been used by many researchers. [52] Although the relationship between myoelectric output and muscle force production is not perfectly linear, EMG analysis is considered to be a semiquantitative measure of the force developed by a muscle. [53-56] The linear envelope method used in our study is one of the least complex and most often used methods of assessing myoelectric responses. [57] A linear envelope consists of a zero offset full-wave rectifier followed by a low-pass filter to average the variation that occurs in the input sEMG signal. The linear-enveloped sEMG signal therefore represents the mathematical average or mean of the rectified input signal; alternatively, it can be thought of as a moving average of EMG activity. Numerous studies have demonstrated sEMG to be a repeatable, stable, and reliable measure of muscle output during various tasks, including isometric MVCs. [58-61] In this study, a series of 3 isometric MVC extension trials were performed to assess erector spinae muscle output. Repeat isometric MVC trials are often used to examine changes in median frequency associated with fatigue. [61, 62] Such trials generally involve protocols with a short rest period between efforts (typically, 2 seconds) combined with long duration for the MVC (typically, 5-15 seconds) over a relatively long period of time (typically, more than 60 seconds). Our MVC protocol was relatively brief in duration and included sufficient time for rest between each trial so that little, if any, muscle fatigue occurred.

We found a significant increase in lumbar muscle sEMG output after SMT and stiffness assessment, whereas no significant pre-versus-post differences in lumbar sEMG output were observed in the 2 control groups. Previous studies investigating the efficacy of SMT in the treatment of patients with LBP have assessed subjective measures of self-reported pain and disability. [1, 4] Pope et al [63] conducted a prospective randomized 3-week trial of SMT, transcutaneous muscle stimulation, massage, and corset in the treatment of subacute LBP. Patients were evaluated once a week through use of questionnaires (including visual analog scale), range of motion, maximum voluntary extension effort, straight leg raising and Biering-Sorensen fatigue testing. After 3 weeks, the manipulation group scored the greatest improvements in flexion and pain. The massage group had the best extension effort and fatigue response, whereas the muscle stimulation group showed the best extension. None of the groups were significantly different with respect to changes in physical outcome measures (range of motion, fatigue, strength, and pain). Other researchers have noted a decrease in quadriceps muscle inhibition and increased knee extensor torque and muscle activation after sacroiliac joint manipulation. [64] The generally favorable objective findings reported by Pope et al [63] and Suter et al64 are consistent with the findings of the present study.

In this study, a prone-position, repeated MVC extension task was used to determine trunk muscle output. MVC trunk extension tasks performed in the prone-lying position have been found to produce the greatest challenge to the erector spinae musculature. [27, 65, 66] The prone posture has also been a commonly used method of quantifying erector spinae strength by means of electromyography.67 Protocols that require subjects to perform 3 isometric MVCs to establish a reference level of maximum isometric trunk extension are often advocated as a normalization procedure and are considered to provide more accurate comparisons between subjects and muscle groups. [65] Quantification of MVC isometric trunk extension, however, includes inherent problems in standardizing a dynamic event such as trunk strength. Potential confounding factors include subject motivation, subject effort, joint angle, velocity of movement, and pain provocation in performance of the effort. [52] Another potential confounding factor is the length of time that the subject lies prone during the testing procedure. In this study, each LBP subject spent approximately 30 minutes lying prone during the MVC, AMCT, PA stiffness, and algometry protocols. Of course there was some variation in the precise amount of time that the subjects in the 3 treatment groups remained in the prone position, because each underwent an AMCT protocol specific to his or her leg check findings. In general, however, our data file records indicated that the 30-minute time interval was fairly consistent across all subjects. In the case of the control group, which did not receive any stiffness or algometry assessments, the rest period was precisely established from the SMT treatment group data file records, which provided us with a precise timeline for the stiffness assessment protocol. Thus, the time spent in the prone position was not a significant variable for any of the groups examined in this study.

Our study design involved assigning 20 subjects with LBP to the sham-SMT and control groups in such a way that age and sex were uniformly stratified with respect to the SMT treatment group. All subjects, however, were given the same verbal directives in terms of the objectives of the study, and no indication was made of patient grouping or the controlled nature of the study to those participating. Subject demographics, health status parameters, and pain tolerance were similar among the 3 groups. Our study design also included a sham-SMT procedure, and the peak forces and peak impulses delivered to the patients during the sham-SMT thrust were substantially less than the peak forces and impulse delivered to the SMT treatment group. Study designs based on a sham-SMT effectively control for any influence associated with the doctor-patient interaction. [68-70] Although the sham-SMT and control trials were performed 18 months after the initial SMT tests, the procedures were performed through use of an identical protocol, in the same clinical setting, and by the same practitioner who uses AMCT exclusively (C.J.C.). Thus, there were no changes in practice patterns associated with the sham-SMT and control subject trials. Moreover, the nature of the sham-SMT (zero force setting) and control (no intervention) trials preclude any confounding effects related to practice patterns. The positive outcomes observed in the treatment group provide additional insight toward understanding the underlying physiologic mechanisms that may be associated with SMT.

From the standpoint of sEMG output, our study appears to be the first to show an acute change in lumbar paraspinal muscle function after SMT. Specifically, our results indicate that an MFMA adjustment in combination with a PA stiffness assessment was associated with a greater than 20% increase in MVC lumbar sEMG amplitude after treatment. We also hypothesized that MVC lumbar sEMG output would not change after sham-SMT and sham-stiffness interventions or no intervention (control). This hypothesis was substantiated by the results obtained from the tests conducted on a group of age- and sex-matched subjects with LBP. These findings suggest that the MVC lumbar sEMG output increases observed in the SMT treatment group resulted directly from the SMT and stiffness assessment interventions. Moreover, we found that the pretreatment/posttreatment MVC lumbar sEMG changes noted for the SMT treatment group were significantly greater than corresponding changes in the sham-SMT and control groups, although a larger group size comparison is necessary to substantiate this finding.

It is important to point out that the positive changes in MVC lumbar sEMG output cannot be directly attributed to the SMT treatment alone, inasmuch as the posttreatment MVC test did not immediately follow the MFMA adjustment. The current study was motivated by preliminary findings obtained from an initial group of 20 subjects with LBP who underwent, in addition to MFMA adjustment, a comprehensive dynamic stiffness assessment protocol through use of a specially designed instrument. [40] Future studies should include more specific treatment modalities to further investigate the effects of SMT-specific interventions on trunk muscle strength.

Currently, numerous basic science research studies are focusing on investigating the mechanical and physiologic basis for SMT. Neurophysiologic models theorize that SMT may stimulate or modulate the somatosensory system and may subsequently evoke neuromuscular reflexes. [33, 34] MFMA SMT has been found to elicit significant neuromuscular reflexes in patients with LBP. [36, 38] Such mechanical and neurophysiologic studies indicate that joint manipulation may have both direct and indirect clinical benefits. Beneficial effects of SMT have been thought to be associated with mechanosensitive afferent stimulation and presynaptic inhibition of nociceptive afferent transmission in the modulation of pain, [71, 72] inhibition of hypertonic muscles, [33, 73] and improved functional ability. [2, 74] Mechano-sensitive and nociceptive afferents have been identified in the lumbar intervertebral discs, [18, 75, 76] zygapophyseal joints, [22, 77, 78] spinal ligaments, [79-81] and the paraspinal musculature [82, 83] in both animal and human studies. In addition, joint stimulation has been found to be intimately related to reflexogenic muscular reactions. [12, 13, 84-86] According to myogenic models that relate abnormal EMG patterns to the experience of pain, lumbar paravertebral muscle activity has been considered to play an important role in chronic LBP. [87, 88] The acute improvements in lumbar sEMG output after MFMA SMT and stiffness assessment suggest that a potential beneficial effect derived from SMT may be acute improvement in muscle functional status. A larger-scale, randomized controlled clinical trial needs to be conducted to further investigate the clinical relevance of these findings.

Considering its inherent limitations, the results of this preliminary study suggest that MFMA or percussive-type SMT applied to the lumbar spine may produce an increase in erector spinae muscle surface electromyography output as measured during maximum voluntary contraction extension tasks. These findings may be related to the ability of mechanical forces to stimulate the somatosensory system, inhibit nociception, improve functional ability of trunk muscles, and/or improve dorsolumbar range of motion. These findings indicate that altered muscle function may be a potential short-term therapeutic effect of these types of clinical interventions.


Considering its inherent limitations, the results of this preliminary study suggest that MFMA or percussive-type SMT applied to the lumbar spine may produce an increase in erector spinae muscle surface electromyography output as measured during maximum voluntary contraction extension tasks. These findings may be related to the ability of mechanical forces to stimulate the somatosensory system, inhibit nociception, improve functional ability of trunk muscles, and/or improve dorsolumbar range of motion. These findings indicate that altered muscle function may be a potential short-term therapeutic effect of these types of clinical interventions.