J Manipulative Physiol Ther. 2011 (Feb); 34 (2): 88–97
Heidi Haavik, PhD, BSc (Chiro), Bernadette Murphy, PhD, DC
New Zealand College of Chiropractic,
Auckland, New Zealand.
OBJECTIVE: The objectives of this study were to investigate whether elbow joint position sense (JPS) accuracy differs between participants with a history of subclinical neck pain (SCNP) and those with no neck complaints and to determine whether adjusting dysfunctional cervical segments in the SCNP group improves their JPS accuracy.
METHOD: Twenty-five SCNP participants and 18 control participants took part in this pre-post experimental study. Elbow JPS was measured using an electrogoniometer (MLTS700, ADInstruments, New Zealand). Participants reproduced a previously presented angle of the elbow joint with their neck in 4 positions: neutral, flexion, rotation, and combined flexion/rotation. The experimental intervention was high-velocity, low-amplitude cervical adjustments, and the control intervention was a 5-minute rest period. Group JPS data were compared, and it was assessed pre and post interventions using 3 parameters: absolute, constant, and variable errors.
RESULTS: At baseline, the control group was significantly better at reproducing the elbow target angle. The SCNP group's absolute error significantly improved after the cervical adjustments when the participants' heads were in the neutral and left-rotation positions. They displayed a significant overall decrease in variable error after the cervical adjustments. The control group participants' JPS accuracy was worse after the control intervention, with a significant overall effect in absolute and variable errors. No other significant effects were detected.
CONCLUSION: These results suggest that asymptomatic people with a history of SCNP have reduced elbow JPS accuracy compared to those with no history of any neck complaints. Furthermore, the results suggest that adjusting dysfunctional cervical segments in people with SCNP can improve their upper limb JPS accuracy.
Key Indexing Terms: Proprioception, Upper Extremity, Manipulation, Spinal, Central Nervous System, Posture, Chiropractic
From the FULL TEXT Article
There is a growing body of evidence demonstrating that adjusting (also known as manipulating) dysfunctional spinal segments can alter central neural function. [1–12] However, it is less certain whether these observed changes in central nervous system (CNS) processing reflect clinically beneficial changes to the individual participants. It has been suggested that these observed changes in sensory processing, sensorimotor integration, and motor control could reflect a mechanism that explains the functional improvements observed after chiropractic care. [6, 10, 12, 13] Our group has proposed that high-velocity, low-amplitude manipulation has a neuromodulatory effect on CNS function.  Furthermore, we have proposed that segments of the spine where the movement is functionally restricted in at least 1 plane may represent an ongoing state of altered afferent input that could induce maladaptive neuroplastic changes. [6, 10, 12, 13] The functional segmental restriction could involve restriction in the coronal plane, such as reduced lateral flexion motion, or could include restriction of appropriate movement in the sagittal plane, such as decreased flexion or extension movement. This functional putative manipulable lesion is known by a variety of terms such as joint dysfunction, fixation, or subluxation. It has been suggested in the literature that the maladaptive neuroplastic changes present in long-term pain conditions rather than the actual pain itself are responsible for the individual sufferer's symptoms and functional disturbances. [14–16] In particular, changes in the way the CNS processes proprioceptive information have been suggested as the most important factor responsible for the clinical presentation of neck pain sufferers. 
One of our previous studies using somatosensory-evoked potentials has shown that adjusting dysfunctional cervical segments of patients without frank neck pain but with a history of some form of subclinical neck pain (SCNP) can alter cortical somatosensory processing and early sensorimotor integration of input from the upper limb.  Subclinical neck pain refers to recurring neck dysfunction such as mild neck pain, ache, and/or stiffness with or without a history of known neck trauma. Individuals with SCNP do not have constant symptoms and have not yet sought treatment of their neck complaint. There is an increasing interest in SCNP in the literature because individuals that fall into this category provide an opportunity to explore neurophysiologic dysfunction without the confounding effect of current pain, which is known to alter sensory processing and motor control.  Furthermore, it is thought that gaining a better understanding of the features characterizing this group may help improve subgrouping of neck pain patients. In addition, it could provide a marker of altered sensory processing that could aid in determining those individuals showing evidence of disordered sensorimotor integration who need treatment to prevent the progression of neck pain into more long-term pain states. 
One possibility for the observed changes in early somatosensory processing at the level of the primary sensory cortex (ie, N20 somatosensory evoked potential [SEP] peak changes) after neck adjustments  (also known in the literature as manipulation) is that this reflects alterations in proprioceptive processing. SEPs are produced by transcutaneous electrical stimulation of a peripheral nerve and are thought to reflect central processing of signals originating from muscle afferents. Information from muscle afferents are known to be extremely important for central proprioceptive processing (for review, see ). Therefore, it is possible that the observed changes in the N20 SEP complex after cervical adjustments  reflect changes in central proprioceptive processing.
It is possible that cervical spinal dysfunction disturbs proprioception from the neck and upper limb and that spinal adjustments improve it. Palmgren et al  demonstrated that chiropractic care can improve head repositioning accuracy, which is an indicator of improved proprioception, suggesting that spinal adjustments can improve spinal proprioception.
A recent study takes the work of Palmgren et al  a step further, as it suggests that cervical spine function can influence upper limb proprioception. Knox and Hodges  demonstrated that changes in head and neck position in a group of participants without any history of neck pain or injury led to reduced accuracy of elbow joint position sense (JPS). The authors of this study discussed how accurate execution of movement depends on the ability of the CNS to integrate somatosensory, vestibular, and visual information regarding the position of the body.  They argued that placing their participants' heads in full flexion and rotation could have led to an overload of the computational capacity of the CNS, thus resulting in increased JPS error.  The same group of researchers also demonstrated that people with whiplash-associated disorder (WAD) are affected by smaller angles of neck rotation than individuals who had no history of WAD,  further suggesting that cervical spine dysfunction leads to reduced accuracy of JPS. Taken together, these studies suggest that spinal function can impact central proprioceptive processing not only of the spine itself, as the study of Palmgren et al  suggests, but also of the upper limb. It is therefore possible that the changes in the N20 SEP peak after spinal adjustments of dysfunctional cervical segments  could reflect such changes in proprioceptive processing of the upper limb.
The aims of the current study were therefore to investigate whether JPS accuracy differs between SCNP participants and those with no history of any neck symptoms or injury and to determine whether manipulating (adjusting) dysfunctional cervical segments in the SCNP group can improve the accuracy of their elbow JPS/
The major findings in this study were that participants with a self-reported history of subclinical neck pain have significantly worse elbow JPS compared to people that have no neck complaints and that a single session of high-velocity, low-amplitude adjustments of dysfunctional cervical joints resulted in a significant improvement of elbow JPS.
The Effect of the Spine on Limb Proprioception
Previous research [20, 28, 34] has demonstrated that both perceived and actual head and neck positions can influence the accuracy of elbow JPS. Thus, the body's internal reference framework appears to be very important for accurate integration of incoming proprioceptive information. It was therefore unexpected that the SCNP participants in the current study did not demonstrate a worsening of JPS because of actual changes in head and neck position; particularly, because Knox et al  have also shown that changes in head and neck positions have a greater effect on elbow JPS when people have had head or neck injuries such as WAD. Of the 25 SCNP participants in the current study, 14 reported a history of previous head or neck injury such as concussion or whiplash injury. However, this discrepancy in findings is most likely due to methodological differences. In the current study, the participants' heads were placed in the various positions before presentation of the target angle and were left in that position for all 6 trials. Knox and Hodges  moved their participants' heads between the presentation of the target angle and before the participants were asked to actively reproduce this target angle. It is tempting to suggest that the distraction of the movement may have influenced the reduced accuracy in elbow joint angle reproduction in that study. However, because the participants in the study of Knox and Hodges were not influenced by the movement distraction control intervention,  this seems an unlikely explanation. It is quite possible because our participants had their heads held in the various positions throughout each set of trials that their CNS may have been able to adapt to this position and more accurately judge their elbow joint position.
The Functional Role of the Spine in Subclinical and Long-term Pain Syndromes
There is evidence in the literature to suggest that muscle impairment occurs early in the history of onset of neck complaints  and that this muscle impairment does not automatically resolve even when neck pain symptoms improve. [35, 36] Some authors have therefore suggested that the deficits in proprioception and motor control rather than the pain itself may be the main factors defining the clinical picture and chronicity of different long-term pain conditions. [14–16] The finding that the SCNP participants have significantly worse JPS accuracy compared to participants with no history of any neck complaint supports this hypothesis. The current study results suggest that deficits in proprioception identified in the SCNP group may be partly due to the presence of the type of spinal dysfunction that chiropractors and other manipulative therapists treat. The cervical adjustment intervention improved the SCNP participants' elbow JPS accuracy to a similar level as that of the control group and to what has previously been reported in the literature in asymptomatic healthy populations with no history of head or neck symptoms or injuries.  This supports the theory that chiropractic care can have a beneficial neuromodulatory effect.  The improvements we observed might be even more impressive in a group with a greater level of pain and disability because some authors [35, 37] have observed larger repositioning errors in persons reporting worse functional disability scores than those with milder problems.
It is also possible that the putative “manipulable lesion,” also known as “vertebral subluxation” or “dysfunctional spinal joint segment,” may represent a state of altered afferent input that may be responsible for ongoing central plastic changes. It is well established that altered afferent input to the CNS leads to changes in CNS functioning. [38–40] Thus, as previously postulated, [6, 10–13] a potential mechanism that could explain how manipulation improves function is that altered afferent feedback from a dysfunctional neck or spine alters the afferent “milieu” into which subsequent afferent feedback from the spine and limbs is received and processed thus leading to altered sensorimotor integration of the afferent input, which is then normalized by high-velocity, low-amplitude adjustments of the dysfunctional areas of the spine.
Limitations and Potential Bias
Another unexpected finding in the current study was that the control participants' elbow joint position accuracy reduced after the control intervention. This methodology has previously been reported to be reliable, [21, 28, 41] at least for the absolute and variable error measurements.  However, the control participants in the current study performed significantly worse after the control intervention. They displayed an overall reduction in absolute error and a greater variable error. However, several of the participants in both groups reported that they felt like their arm was “going to sleep” because of maintaining the supine position with their arm externally rotated and abducted for the length of time the data recording session took. This has not been reported by previous studies using the same participant positioning as used in the current study. [20, 34]
However, our experiment took longer because we carried out pre and post measures in the same session, and this alone could explain the differences we observed. Juul-Kristensen et al  used a seated participant position in their test-retest reliability study, which could explain the better reliability findings for absolute and variable errors. It is therefore possible that the participant positioning in the current study has led to upper limb sensory disturbances because of position-related compression effects on the brachial plexus that caused the worsening of the absolute and variable errors in the control participants. If this is the case, the improvements seen in the SCNP group after the adjustment intervention are all the more impressive because several of the SCNP participants also complained that maintaining the required position for the duration of the data collection procedure was affecting their upper limb.
Another possibility is that the control participants worsening JPS accuracy after the control intervention was because of a boredom effect and that the improvements seen in the SCNP group was because of a placebo effect or that both groups participated with a different degree of effort (ie, the avis effect). It is very difficult to avoid these potential biases because sham adjustments (ie, sham spinal manipulation) are near impossible to perform. This needs to be addressed in future studies. Furthermore, the potential biases, although they must be considered, cannot explain the reports from the participants that the position was causing their right arms to go to sleep (described as mild numbness and tingling sensations).
The results of this study suggest that asymptomatic people with a history of recurring neck pain, stiffness, or ache have reduced elbow JPS accuracy compared to those with no history of any neck complaints. Furthermore, the results suggest that even a single session of adjusting dysfunctional cervical segments in people with subclinical neck pain can improve their upper limb JPS accuracy.