From: J Manipulative Physiol Ther. 2010 March – April;33(3):168-177.
Chronic neck pain is a common problem in modern, industrialized countries. It has been estimated that 67% of people will experience neck pain at some point in their lives. A proportion of these individuals with neck pain do not experience complete resolution of their pain and disability, which can turn into a more complex chronic pain syndrome. What is not well understood is what causes neck pain to become chronic. An emerging school of thought in the mechanism of chronicity in nonspecific neck pain is that it is related to abnormal muscle recruitment patterns, which may put the spine at greater risk of further injury. Impaired neuromuscular function in patients with chronic neck pain is becoming increasingly recognized, most notably, an impaired ability to activate the neck flexor muscles during rapid limb movements and an impaired ability to relax the neck extensor muscles.
The solution proposed in previous research has been based on the idea of using specific exercise strategies to improve these impaired neuromuscular patterns. Recent research indicates that both exercise and chiropractic care involving spinal manipulation may also be able to improve these impaired neuromuscular patterns. Chiropractic techniques appear to be able to help normalize altered patterns of muscle recruitment and sequencing observed in the presence of musculoskeletal impairments and pain.
Contemporary research into the pathogenesis of nonspecific neck pain relates to the manifestation of abnormal muscle recruitment patterns. Impaired neuromuscular function in patients with chronic neck pain is becoming increasingly recognized, most notably the impaired activation of the neck flexor muscles during rapid upper limb movement. An additional measure that may be used for impaired neuromuscular function is the cervical flexion-relaxation response, a measure of the ability to relax the cervical extensors at full forward flexion. There is a lack of evidence for how commonly used interventions for chronic neck pain, such as spinal manipulation or exercise, may change these measures of impaired neuromuscular function in the neck.
The flexion-relaxation response, commonly measured in the posterior kinetic chain muscles (such as the erector spinae, hamstrings), is the electrical silence that is observed during full forward flexion of the trunk or neck as the passive structures of the spine maintain stability. Individuals with either chronic neck or back pain have been shown to exhibit heightened muscle activity at full flexion of the neck and trunk, respectively. Recent results suggest that for the lumbar spine, the flexion-relaxation response may be an important marker of neuromuscular impairment, which can show improvement with an appropriate intervention. After an exercise intervention for individuals with chronic low back pain, it was shown that changes in the flexion-relaxation response explained 38% of the improvement in self-reported disability. This change seems to be predominantly manifested as improved relaxation at full trunk flexion. Although the cervical flexion-relaxation response is a reliable measure able to discriminate between patients with and without chronic neck pain, it is not known if this response will change after treatment interventions.
Similar to the deep abdominal muscles, the cervical flexor muscles have a pattern of activating within 50 milliseconds before the onset of activity in the deltoid muscles during rapid upper limb movements in healthy subjects. Falla et al (2004) showed impaired feed-forward activation in people with chronic neck pain. When subjects with a history of neck pain performed arm flexion, the activation onsets of deep cervical flexors ipsilateral to the sight of arm movement as well as the contralateral (right) sternocleidomastoid and anterior scalene muscles were significantly delayed compared with the relative latencies for the control group.
A recent review by Vernon et al found moderate to high quality evidence that manipulation at 6, 12, and up to 104 weeks was an effective treatment for neck pain not due to whiplash with effect sizes from 0.56 to 3.2, most of which would be considered large. These effect sizes were maintained up to 12 weeks posttreatment. Only 2 studies had long-term data but still showed large effect sizes for up to 104 weeks. A Cochrane review found a strong benefit of manipulation and/or mobilization combined with exercise vs waiting list controls for pain reduction, improvement in function and global perceived effort. What has not been assessed in most neck pain intervention studies is whether the treatment is able to affect neuromuscular function. It may be that manipulation is able to affect one type of neuromuscular impairment and another specific exercise, and that the combined treatment effect may be synergistic. It may also be that improved neuromuscular function may be a marker of an effective treatment.
The purpose of this study was to evaluate manipulation-based chiropractic care combined with strengthening exercises as compared with strengthening exercises only for the treatment of chronic neck pain, with the inclusion of the cervical flexion-relaxation response and cervical feed-forward activation latencies as measures of neuromuscular activation.
The main hypotheses of this study were (1) that improvements in functional capacity would be greatest for patients who received the combination of chiropractic care and exercise compared with exercise only and (2) that these improvements would be reflected by changes in neuromuscular deficits.
The major finding in this study was that chiropractic care combined with 8 weeks of exercise and exercise alone are both effective at reducing perceived levels of functional neck disability and pain. The null hypothesis was proven in this study because there were no significant differences between the 2 groups. The effective size calculations for the differences between the 2 groups was .293, which is a small effect,28 and the sample size calculations indicated that 145 subjects per group would be needed to show a difference between the 2 groups based on the NDI. Population scores suggest that both groups had mild disability at baseline, and although a decrease in score was observed in both groups, the classification did not change after the intervention. It may be because both groups were in the “mild category,” and it was a pilot study that the study lacked sufficient power to show a difference between treatments. There was also a lot of variability in the degree of change between individual subjects as shown by the large SDs, which had the effect of decreasing the effective size and increasing the estimated sample size required to show a difference between the 2 groups. Another important consideration is whether the small difference between the 2 groups has enough clinical relevance to justify the costs of an RCT involving 290 participants.
There was a decrease for both groups in current pain scores and “worst pain,” which was significant overall, but again, because of the large degree of variability between the groups, the effect sizes were small for both, and sample size estimates indicated that 88 subjects per group would be needed to show a difference between treatments for “pain now” and 82 per group for “worst pain.”
Improvements in the flexion-relaxation results would be indicated by an increased flexion-relaxation response, indicating improved relaxation (decreased myoelectric activity at full flexion). Both groups would be considered initially impaired based on previous work in the lumbar spine. The flexion-relaxation response actually worsened slightly for the exercise group and improved minimally for the manipulation group. This result contrasts with previous work, which showed that an 8-week exercise intervention was able to improve the flexion-relaxation response in the lumbar spine. The fact that the manipulation group group that had manipulation before exercise improved slightly is in keeping with previous work in the lumbar spine. This trend toward improvement in the manipulation group group as compared with the exercise group group is in keeping with the authors initial hypothesis and suggests spinal manipulation before exercise may help to normalize neuromuscular function and enable people to better cope with the demands of exercise. Given the reasonably low number of participants required to determine whether the 2 forms of treatment have differential effects on this outcome measure, the authors could recommend a future RCT with 40 participants per group to allow for dropouts.
Improvements in feed-forward responses would be indicated by faster onset times in the cervical flexor muscles in relation to deltoid onset times (within 50 milliseconds). Previous work by Falla et al indicated that chronic neck pain patients exhibit delayed activation of the cervical flexor muscles. The baseline results for feed-forward activation were not impaired in the chronic neck pain patients who participated in this trial when compared with Falla’s results. This could explain the lack of finding for the anterior scalene muscles in the manipulation group group compared with Marshall and Murphy’s study where only subjects who showed evidence of delayed feed-forward activation received spinal manipulation therapy. The authors do not think the feed-forward changes reported in this study should be used for sample size calculations, because the participants did not actually show delayed activation before the treatments.
An important consideration in interpreting our results is that 8 weeks may not be enough time to show the full range of improvement in neuromuscular measures. A more recent article by Marshall and Murphy for the low back has indicated that at a 9-month follow-up feed-forward activation had improved in a group of chronic low back pain patients even though these changes were not present after 12 weeks of exercise.
This pilot study showed that both exercise and exercise combined with manipulation can improve pain and disability in people with long-term neck pain. The study indicates that the flexion-relaxation response changes had an effective size of .636, and 32 subjects per group would be needed to show a difference between the 2 treatments with an a of .05 and a power of 0.8.
