Gregory Cramer, DC, PhD, Brian Budgell, DC, PhD, Charles Henderson, DC, PhD,
Partap Khalsa, DC, PhD, Joel Pickar, DC, PhD
Department of Research,
National University of Health Sciences,
Lombard, Ill. 60148, USA.
OBJECTIVE: The objectives of this white paper are to review and summarize the basic science literature relevant to spinal fixation (subluxation) and spinal adjusting procedures and to make specific recommendations for future research.
METHODS: PubMed, CINAHL, ICL, OSTMED, and MANTIS databases were searched by a multidisciplinary team for reports of basic science research (since 1995) related to spinal fixation (subluxation) and spinal adjusting (spinal manipulation). In addition, hand searches of the reference sections of studies judged to be important by the authors were also obtained. Each author used key words they determined to be most important to their field in designing their individual search strategy. Both animal and human studies were included in the literature searches, summaries, and recommendations for future research produced in this project.
DISCUSSION: The following topic areas were identified: anatomy, biomechanics, somatic nervous system, animal models, immune system, and human studies related to the autonomic nervous system. A relevant summary of each topic area and specific recommendations for future research in each area were the primary objectives of this project.
CONCLUSIONS: The summaries of the literature for the 6 topic sections (anatomy, biomechanics, somatic nervous system, animal models, immune system, and human studies related to the autonomic nervous system) indicated that a significant body of basic science research evaluating chiropractic spinal adjusting has been completed and published since the 1997 basic science white paper. Much more basic science research in these fields needs to be accomplished, and the recommendations at the end of each topic section should help researchers, funding agencies, and other decision makers develop specific research priorities.
From the FULL TEXT Article:
Chiropractors use a variety of diagnostic procedures and therapies to treat a wide array of disorders. The mode of treatment that is common to all chiropractors and most frequently used is spinal adjusting (manipulation). As a result, since the 1997 basic science white paper,  there has been significant research activity evaluating the mechanisms of spinal fixation and spinal adjusting. This white paper is focused on that activity. The objectives of this article are:
(1) to review and summarize the basic science literature relevant to spinal fixation (subluxation) and spinal adjusting procedures since the publication of the last white paper in 1997, and
(2) to make specific recommendations for future research.
A. Anatomical Research
The topic of anatomical research will be divided into studies that relate specifically to SM (“Studies Related to SM”) and those that add detail to the understanding of the spine (“Studies Adding to the Knowledge Base of the Spine”).
Dysfunction of the spine that is treated by chiropractors has been described as the vertebral subluxation complex. This complex, as described by Lantz  and Rosner,  has several components. These components include myologic, connective, vascular, neurologic, and lymphatic tissue involvement. Many hypothesize that a fundamental component of the vertebral subluxation complex is the development of adhesions in the zygapophysial joints (Z joints) after hypomobility of these structures. [4-6] Spinal adjusting of the lumbar region is thought to separate the articular surfaces of the Z joints. [7-9] This “gapping” is theoretically the action that “breaks up” adhesions. Elimination of adhesions would allow the Z joints to become more mobile, thus helping the motion segment (2 adjacent vertebrae and the ligamentous structures connecting them) to reestablish a physiologic range of motion (ROM).  Although the idea that gapping the Z joints would occur during a SM and that this action would break up adhesions seems logical to those who have performed these procedures, no anatomical or physiologic evidence existed  to show that gapping of the Z joints actually occurred, or that adhesions in the Z joints could potentially be broken until relatively recently. [10, 11] Cramer et al [10, 11] found that in healthy volunteers, the lumbar Z joints did gap during both side-posture positioning and during chiropractic adjusting and the joints gapped significantly more during the latter procedure. However, no reports have been published by any group regarding gapping of the Z joints in the low back pain population.
Degeneration of the z joints after hypomobility
Paris  has reportedly identified adhesions in the Z joints after hypomobility. More recently, degenerative changes have been identified in the Z joints of rats after induced hypomobility.  See the section on Animal Models for further details.
Advances in magnetic resonance imaging relevant to the spine and spinal adjusting
Functional magnetic resonance imaging (MRI) evaluation of the spinal cord maps MRI signal changes following a specific stimulus designed to change neural activity. This procedure may become fundamental in the future workup of spinal cord injury  and may provide intriguing possibilities to the assessment of the spinal cord following procedures such as spinal adjusting. Diffusion and perfusion MRI provides information related to the structure and function of tissue at a microscopic level and will play a more prominent role in future neurovascular imaging.  As MRI at the molecular level becomes possible, opportunities for advances in research (and clinical practice) increase. 
Research is also being done involving morphometry of the spine by means of MRI.  Morphometric measurements allow for an increased ability to study the structures influenced by chiropractic adjusting. [11, 17] Of related interest is that direct oblique cervical MRI provides more accurate assessment of all of the borders of the cervical intervertebral foramens (IVFs) than standard sagittal MRI. 
Between 15% and 40% of chronic low back pain is related to the Z joints.  The Z joint capsule receives a significant sensory innervation, [19, 20] much of which is probably related to nociception, that is, signaling potential or real tissue damage. The medial branches of the posterior primary divisions innervating a Z joint terminate as 1 of 3 types of sensory receptors:
free nerve endings (nociceptive),
complex unencapsulated nerve endings, and
encapsulated nerve endings.
The free nerve endings are associated with nociception. The ultrastructure of these receptors has been described. [20, 21]
The Z joint capsules throughout the vertebral column are thought to do little to limit motion.  However, the capsules probably help to stabilize the Z joints during motions.  The gross and microscopic anatomy of the Z joint capsules has been described in detail. [23, 24] See the sections on Biomechanics and Somatic Nervous System for further details of current research related to the Z joints and spinal adjusting.
Ligaments of the spine
Nerve tracing techniques indicate that stretching of spinal ligaments results in “a barrage of sensory feedback from several spinal cord levels on both sides of the spinal cord.”  The sensory information has been found to ascend to many higher (cortical) centers. Such findings provide provocative evidence that the spinal ligaments, along with the Z joint capsules, and the small muscles of the spine (interspinales, intertransversarii, and transversospinalis muscles) play an important role in mechanisms related to spinal proprioception (joint position sense) and may play a role in the neural activity related to spinal adjusting. 
Recent work has also been done assessing the structure of spinal ligaments. The attachment sites and dimensions of the anterior and posterior longitudinal ligaments [27, 28] and the innervation and gross and light microscopic structure of the ligamenta flava [29, 30] and iliolumbar ligaments  have been studied in detail.
The long posterior sacroiliac ligament may be important in transmitting loads from the lower extremity to the spine.  The strongest fibers course from the posterior superior iliac spine to the sacrotuberous ligament, and many important structures attach to this band, including the aponeurotic attachments of the common origin of the erector spinae muscle. The ligament is tensed during counternutation of the sacrum and slackened during nutation.  These findings are considered to be important by those involved with the study of the “kinetic chain concept” of load transmission from the lower extremity to the spine.
The intervertebral disk and intervertebral disk degeneration
Many relevant studies on the biology of the intervertebral disks (IVDs) have been completed in recent years. Disk degeneration is characterized by loss of fluid pressure, disruption or breakdown of collagen and proteoglycans, and sclerosis of the cartilaginous end plate and the adjacent subchondral bone. All of these hallmark signs of IVD degeneration can also occur as part of the normal aging process of the IVD. For these reasons, disk degeneration and normal aging of the disk are frequently discussed interchangeably, [33, 34] although the biochemical processes may be distinct. The IVD seems to age differently from other tissues, probably because of its lack of a blood supply, and the degenerative process may begin as early as 20 years of age (earlier in some cases).  In fact, certain teenagers may experience back pain because of IVD degeneration.  There is an extremely wide variation in aging and degeneration of the IVD. Some individuals in their 70s have disks of equivalent health to some in their 30s. The aging and degenerative stages of the IVD from prenatal development through the ninth decade of life have been worked out in considerable detail at the gross and light microscopic levels. [34-40] Calcification of the IVD during the aging process is much more common than was once thought, being found in 58.3% of subjects at autopsy. Such calcification is “significantly underestimated” by conventional radiography. 
Several conditions promote or even possibly initiate disk degeneration. These include traumatic Schmorl’s node formation, advanced aortic atherosclerosis,  and possibly, nicotine consumption.  The biochemistry of IVD degeneration is also being elucidated. In this regard, extruded nucleus pulposus has been found to spontaneously produce increased amounts of many chemokines that not only initiate a series of events that decrease the size of the IVD bulge but also result in IVD degeneration. [26, 35, 44-47]
Intervertebral disk protrusion
The normal mechanics of the IVD continue to be investigated. [33, 48-51] In addition, the mechanisms involved in IVD protrusion and failure have been studied in detail [35, 39, 52-62], as well as the effects of changing intradiscal pressure. [63, 64, 65] In addition, a set of terms to be used when describing bulging of the IVD was established by the International Society for the Study of the Lumbar Spine.  This terminology included disk bulge, protrusion (tearing of some inner layers of the anulus fibrosus with the nucleus extending into the radial tear), extrusion (tearing of all layers of the anulus fibrosus allowing nuclear material to enter the vertebral canal), and sequestration (a piece of extruded nucleus breaks off of the host IVD). Much recent research related to IVD protrusion in the cervical and lumbar regions (protrusion in the thoracic region has not been studied as extensively) has found that IVD protrusion is a very dynamic process and that after approximately 1 to 3 weeks, IVD protrusion will usually begin a 2-month to 1-year process of resolution, resulting in significant resorption and, from a patient’s standpoint (ie, pain), often complete remission of signs and symptoms. [33, 67-69] In fact, histologic evidence of resorption of sequestered nuclei pulposi has been found, [70. 71], and shrinkage of protruded nuclei pulposi has been seen on both computed tomography and MRI.  This provides hope to patients with protruded IVDs and for those using conservative methods to treat this condition. Adenovirus-mediated transfer of genes and the resultant production of therapeutic growth factors are being investigated as a means to further study the biology of the IVD and the potential for treatment of disk degeneration ; however, the low vascularity of the adult IVD may preclude the effective use of gene therapy in IVD disease.  Two published studies have shown that the inhibition of tumor necrosis factor–a (TNF-a) (extruded nucleus pulposus contains high levels of TNF-a) by a monoclonal antibody (Remicade [infliximab], Centocor, Inc., Horsham, Pa) is successful in alleviating sciatica. [74, 75] Finally, the mechanisms of radicular pain continue to be studied. [41, 76-78]
Innervation of IVDs
The significant innervation of the IVDs continues to be investigated in detail. [79-82] Degenerated disks have been found to receive increased innervation by sensory fibers conducting nociception.  The added innervation seems to be stimulated by Schwann cells of the nerves innervating the outer aspect of the anulus fibrosus.  Consequently, injured or degenerated disks are likely to be more sensitive to pain than normal disks.
Unique characteristics of the cervical IVDs
The cervical IVDs have been found to differ significantly from the lumbar disks. Rather than being made up of many lamellae, the anulus fibrosus of each cervical disk is composed of a single, crescent-shaped piece of fibrocartilage that is thick anteriorly and becomes very narrow laterally and posteriorly. 
Range of motion studies
Noteworthy studies measuring both the ranges of motion in various regions of the spine (eg, cervical region) and the motions between individual vertebrae continue. The latter activity has led to studies attempting to better understand the concept of coupled motions in the spine. Finally, significant findings related specifically to motions in the sacroiliac joints have also been published in recent years. These findings are summarized next.
Although ROMs (eg, cervical ROMs) can be measured reliably,  measurements made on different days of the same individual can vary considerably.  Coupled motion (eg, rotation of vertebrae during lateral flexion) of spinal segments continues to be actively studied. Current investigators are finding that:
(2) all spinal motions are coupled motions; and
(3) coupling differs from 1 motion segment to the next.
Furthermore, consensus has not been reached on many of these motion patterns. 
The full ROM of the sacroiliac joint is not expressed until the extremes of hip motion are reached, moving an average of 7.5° (range, 3°-17°) in the sagittal plane during full flexion and extension of the hips.  Motions as high as 22° to 36° have been reported in preteen and early teenage children.  Contraction of the left and right transversus abdominis muscles increases stiffness of the sacroiliac joint, thus potentially reducing sprains of the ligaments that protect it. 
Morphometry means “measurement of an organism or its parts.” The past decade has seen many morphometric studies of various spinal structures. These studies allow for more accurate biomechanical and computer modeling (finite element analysis) studies to be performed and also allow for more accurate patient treatment protocols (surgical and manipulative) to be designed. Table 1 shows many of the morphometric studies performed since 1995, the region of the spine investigated, and the specific anatomical structure analyzed.
Certain anatomical findings can best be discussed with each spinal region. The following sections describe anatomical findings of particular significance in the cervical, thoracic, lumbar, and sacroiliac regions. Each of the topics discussed is related to an active area of research.
Connective tissue attachments to the posterior aspect of the spinal dura arising from the foramen magnum, posterior arch of C1, the spinous process of C2,  the rectus capitis posterior minor muscle, [115-117] the ligamentum nuchae, [118, 119] and the ligamenta flava between C1-C2 and C6-C7 [120, 121] have been described. These attachments may hold the dura mater posteriorly during cervical extension (to prevent buckling of the dura mater into the spinal cord) and flexion (to prevent the dura from moving forward and compressing the cord). Some authors have speculated that increased tension of the cervical paraspinal muscles may traction the connection between the rectus capitis posterior minor muscle and the dura, leading to headaches secondary to dural tension.  Others have proposed that tearing of these connective tissue attachments during the flexion component of flexion-extension (whiplash) type of injuries or other trauma to the cervical region could lead to buckling of the dura mater into the cervical segments of the spinal cord. Such dural buckling could conceivably result in the chronic neck pain, headaches, disorders of balance, and signs and symptoms of cervical myelopathy experienced by some patients who have had trauma to the cervical region.  The experimental work of Shinomiya et al [120, 121] lends support to these theories. In addition, homologous attachments in the lumbar region, called meningovertebral ligaments, have been shown to traction the dura mater and the related nerve roots after IVD protrusion.
The structure of the vertebral artery continues to be studied. The tortuosity of the vertebral artery can occasionally increase with severe, multilevel IVD degeneration.  Haynes et al  found that usually, there is no compression or stenosis of the vertebral artery with atlantoaxial rotation. Li et al  found that extreme extension and extension with rotation resulted in decreased flow in both vertebral arteries. Licht et al  found a decrease in flow in the vertebral artery contralateral to rotation and, for the first time, documented an increase in flow on the ipsilateral side of rotation. Mitchell et al  found a decrease in flow through both the left and right vertebral arteries (more in the contralateral vessel) with maximal rotation, especially in those arteries with underlying pathology (eg, atherosclerosis). Therefore, maximal rotation and extension seem to decrease flow through the vertebral arteries, but submaximal rotation seems to have less of an effect.
The cause of idiopathic scoliosis remains unknown, but it is thought to be the result of many factors. A genetic component [127, 128] to the disorder is likely with secondary factors that include a decrease in melatonin production [129-131] and a related increase in circulating levels of the hormone calmodulin.  Changes in skeletal muscle,  connective tissue [46, 47] bone density, rib distortion,  decreased height of the posterior vertebral arch,  asymmetry of the neurocentral synchondrosis,  and the relatively common finding (up to 26%) of syrinx formation and other neuroanatomical abnormalities in the spinal cord [135, 136] are probably the result of the altered biomechanics of the spine and spinal cord that occur with spinal curvatures.  In addition, increasing evidence exists supporting the theory that the primary disorder (probably related to genetic influences) is the involvement of high (cortical) brain centers involved in processing vestibular information. [137-139] An interesting study by D’Attilio et al  may indicate that there could be more than 1 subpopulation of patients with idiopathic scoliosis. These investigators showed that the alignment of the spinal column may be strongly influenced by dental occlusion and temporomandibular joint disturbance. These investigators reported that all 15 rats in an experimental group with an induced malocclusion developed thoracolumbar scoliosis within 1 week after the intervention. The scoliotic curvature was then completely resolved in 83% of these rats 1 week after correction of the malocclusion. None of the 15 untreated control rats in their study developed scoliosis. They suggested that an anatomical and functional relationship between the stomatognathic apparatus and the spinal column could explain their observations. They noted the presence of convergent sensory inputs to the craniocervical cord from stomatognathic and cervical spine structures and posited that a consequential tilt of the first cervical vertebra (C1) could affect the alignment of adjacent vertebrae, destabilizing the vertical alignment of the spine.
Lumbar intervertebral foramina
Transforaminal ligaments of the IVF can be identified on MRI, with a positive predictive value of 86.7%.  These ligaments are present in approximately 60.0% of lumbar IVFs (66.7% of L5-S1 IVFs)  and have been implicated as both a cause of low back pain and nerve root entrapment. [142-146] These structures can be quite sturdy (especially at the L5-S1 region) and can calcify.  They have been found to decrease the dimensions of the compartment transmitting the anterior primary division of the spinal nerve by 31.5%.  Limiting the size of the compartment in this way may, at times, contribute to the incidence of neurologic symptoms in the region, especially after trauma or secondary to degenerative arthritic changes in the region of the IVF. 
Acquired accessory sacroiliac joints frequently (19.1%) form within the posterior (fibrous) portion of the joint. These accessory joints are more common in obese and older individuals and are also associated with other signs of degeneration and periodic low back pain. 
Continue to investigate the effects of spinal adjusting on the tissues of the spine and other organ systems (see section on Autonomic Nervous System) in various disease states (eg, gapping studies of the Z joints).
Continue to evaluate the causes of hypomobility of vertebral segments in the general population, under what conditions such hypomobility is maintained, and continue to characterize the changes of the tissues of the spine after hypomobility (and possibly hypermobility) when normal activity is reestablished (ie, normal forces to the spine are reestablished) and also after spinal adjusting is added in an attempt to help reestablish normal forces and movement to hypomobile tissues. A combination of human and animal studies will be needed to achieve this recommendation.
Evaluate the effects of spinal manipulative procedures on pain of radicular origin and on radiating pain.
Further evaluation at the basic science level of the issue of vertebral and basilar artery iatrogenic pathology is warranted.
Conduct descriptive studies to clarify regional differences (ie, between the cervical, thoracic, and lumbar regions) in the anatomy of the vertebral column and related spinal tissues.
Conduct studies evaluating the normal development of all spinal tissues from embryogenesis through the mid 20s.
Conduct studies further evaluating the aging spine (fourth through ninth decades).
All of the recommendations above should be carried out in both human and animal studies at the gross (imaging and postmortem studies), light microscopic (biopsy specimens, eg, those obtained during surgery, cadaveric studies), and electron microscopic (biopsy specimens and postmortem studies) levels.
Biomechanics is the study of the effects of loads applied to biologic cells, tissues or systems. Biomechanics has its origins from Galileo’s studies of mechanics in general and his creation of the term mechanics as a subtitle of his book “Two New Sciences” (1638) to refer to force, displacement, and strength of materials. Arguably, the “father of biomechanics” is Giovanni Alfonso Borelli, who published in “De Motu Animalium” (1681) the principles of muscle movements based on statics and dynamics. However, the word “biology” and its concept as the study of living organisms did not occur until 1802 when the German naturalist, Gottfried Reinhold Treviranus, published his first volume “Biologie; oder die Philosophie der lebenden Natur”. To rigorously understand SM and its effects requires an understanding of the principles of biomechanics.
A. Manipulation Forces
Since the publication of the first white paper in 1997, there have been several important studies that have further clarified the loads that are applied during SM, and especially during high-velocity, low-amplitude (HVLA) SM. Triano and Shultz  measured the total force that was transmitted through the body during a side-lying lumbar or lumbosacral HVLA SM. The transmitted forces were similar to the applied forces for their temporal history, but the transmitted forces and moments were shown to vary substantially based on patient positioning. Herzog et al  measured the force distribution during thoracic HVLA SM and concluded that there was an important distinction between the total and effective applied forces, with the latter being much smaller than the total applied force. They found that the total peak force was being applied over a mean contact area of 34.8 cm2, but for the thoracic spine, the physiologic contact area of the transverse processes was only 0.25 cm2 (less than 1/100 of the total contact region). Hence, most of the total peak force was being applied to soft tissues (eg, skin, muscle, and fat), and only a small portion (~5 N) was being applied to the transverse process. A similar finding was reported by Kirstukas and Backman,  who reported that the “intense contact area” was on the order of 10 cm2 during thoracic HVLA SM. Clearly, the effective applied force during HVLA SM in general will vary based on the contact area of the manipulator’s hand and the aspect of the vertebra, but in general, the effective force will be less, and sometimes substantially so, than the applied force.
The 3-dimensional force applied during HVLA SM of the cervical, thoracic, and sacroiliac regions has now been measured.  The 3-dimensional data showed that forces in plane with the back (ie, Fx and Fy or shearing forces) always occurred during the SM, which was dominated by the normal (ie, Fz or perpendicular) applied force. The shearing forces were considerable in magnitude, ranging from a low mean of 15% (at T4-5) to a high of 29% (at sacroiliac) of the peak Fz force. As has been previously reported by others,  there was a consistent drop in the preload force magnitude just before the impulse portion of the HVLA SM, which is speculated to be due to a “countermovement” affect.
The role of sex in developing force magnitude has been investigated.  The only previous report that compared male and female manipulators found no significant differences during HVLA SM using a patient simulator.  Forand et al  used an experienced matched group (range, 1-24.5 year of experience) of female and male chiropractors (14 per group) and found that there were no significant differences between sexes in thoracic HVLA SM forces. The one exception was that, in the lower thoracic spine, men applied significantly greater preload than did women.
Another type of SM is mobilization or low-velocity low-amplitude SM, which is commonly used by physical therapists as well as other health professionals, including chiropractors. The general approach is to apply an increasing force over 5 to 10 seconds to determine the “end feel,” and then so calibrated, to apply a slow oscillation (~1 Hz for 10 seconds) about a mean graded force (I-IV arbitrary scale), which is less than the end feel.  Using an instrumented mobilization table, it was found that there was considerable variation in the force magnitudes used by experienced therapists for end feel, as well as grades I-IV mobilizations of L3 vertebra in healthy subjects.  When comparing treatment of younger vs older healthy subjects, it was found that, although mean forces were similar, smaller amplitudes and higher frequency of oscillations were used with older patients.  In a study of patients with nonspecific low back pain, there was considerable variation in the magnitudes of forces used, but the variation was strongly influenced not by the patient’s severity of complaint but by the physical therapist’s training. 
Our understanding of the kinematics of SM has been increased by 2 different types of investigations. First, Keller et al [160-162], have published 2 studies using mechanical force, manually assisted, short-lever SM (ie, Activator [Activator Methods International, Phoenix, Ariz] or very HVLA [VHVLA]) in vivo on patients undergoing lumbar surgery. Using forces ranging from 30 N (lowest setting) to 150 N (maximum setting) on the adjusting instrument, the vertebra where the force was applied had peak displacements of approximately 0.5 mm occurring within 10 milliseconds. Intersegmental displacements occurred of similar magnitudes but with large oscillations lasting 2 to 3 times longer (ie, 20-30 milliseconds), but all oscillations appeared to have damped out within 100 milliseconds. In a second in vivo study, they found that the vertebral displacements due to the Activator instrument were slightly larger (mean, ~0.62 mm) and did not vary significantly, depending on whether the instrument was positioned over the spinous process or facet joint (left or right). 
0 Second, using intact cadaveric human lumbar spine specimens, Ianuzzi and Khalsa  simulated lumbar HVLA SM while measuring vertebra kinematics and facet joint capsule strain. During simulated HVLA SM, the applied loads were within the range measured during in vivo HVLA SM. Vertebral translations occurred primarily in the direction of the applied load and were similar in magnitude (on order of 1-2 mm) regardless of manipulation site. Vertebral rotations (on order of 1°-3°) and facet joint capsule strain magnitudes (on order of 5%) during simulated HVLA SM were within the range that occurred during physiologic motions.  At a given facet joint capsule, distal manipulations induced capsule strains similar in magnitude to those that occurred when the manipulation was applied proximally.
The mobility of lumbar vertebrae in healthy volunteers during mobilization has been assessed using dynamic MRI. Powers et al  found that applying a 10-second grade IV posterior to anterior (PA) mobilization (~100 N force) at the spinous process of a lumbar vertebra produced an extension of the vertebra ranging from a mean of 1.2° at L2 to 3.0° at L5. Using plain film radiographs, Lee and Evans  found similar displacements for a 150 N PA mobilization at L4. Kulig et al,  also using dynamic MRI, found that applying a PA mobilization induced intersegmental motion in all lumbar vertebrae, caudal and cranial, to the site of applied force. This is consistent with the findings of Ianuzzi and Khalsa  who also found that simulated HVLA SM at a single vertebra induced motion in all other lumbar vertebrae. Thus, it is not possible to move only a single vertebrae with SM (high or low velocity) because the spine is a linked and coupled structure.
HVLA SM is commonly associated with a “cracking” sound, which has previously been shown to be associated with a cavitation phenomenon in the facet joints. [169, 170] In healthy volunteers, Ross et al  found that single HVLA SM were typically associated with multiple cavitations (ranging from 2 to 6), which were from nearby vertebrae. This was consistent with the findings of Beffa and Mathews,  who found no significant relationship between the location of the cavitation and HVLA SM of the L5 or sacroiliac joint in asymptomatic volunteers. There is some question as to whether HVLA SM can actually induce motion into the sacroiliac joint, as Tullberg et al,  using stereo radiography, were unable to measure any significant motion of the sacrum relative to the ilium after a combination of HVLA and mobilization SM in patients with “subluxated” sacroiliac joints. Furthermore, Flynn et al  found no association between an “audible pop” and improvement in ROM, pain, or disability in patients with nonradicular low back pain.
An intriguing question has begun to be answered relating to whether changes in intersegmental stiffness can be discerned using clinically available tools. Colloca et al  measured intersegmental impedance (dynamic stiffness) of lumbar vertebrae and correlated it with characteristics of vertebral height and IVD height measured from plain film radiographs. They found that there was a correlation between decreased disk height at L5-S1 and increased dynamic stiffness at the same segment. These findings were analogous to those of Kaigle et al  who, using a porcine model, also observed increased spine dynamic stiffness associated with degenerated disks, compared with normal controls.
Using ultrasound indentation, another noninvasive approach, Kawchuk et al  also found that IVD degeneration in a porcine model resulted in decreased indentation for the same applied load. This is an analogous metric as spine stiffness. The use of ultrasound indentation in this animal model had high sensitivity (75.0%), specificity (83.3%), and accuracy (77.1%), compared with other approaches (arthroscopy, MRI, and plain film radiography).
Two biomechanics studies have been performed to examine the effects of fixation (ie, a hypomobile subluxation) of the lumbar spine. Cramer et al  used a rat model of fixation in the lumbar spine by externally fixating the spinous processes of L4-L6 for up to 8 weeks. A principal finding due to the fixation was the development of osteophytes and degenerative articular changes of the facet joints within a few weeks. Reversal of some of the degeneration was observed for joints that were fixated for a short term (~1 week), but after 4 weeks, no reversal was observed. Little et al  simulated a hypomobile subluxation in intact, cadaveric human lumbar spine specimens by screwing a plate into the left anterior aspect of the L4 and L5 vertebral bodies. During physiologic motions of the fixated spine specimens for flexion, extension, and lateral bending, the motions at L4-5 were significantly decreased, whereas below and above that level, intersegmental motions were significantly increased. Correspondingly, the plane strains of the facet joint capsules were significantly decreased and increased at and above/below the site of fixation, respectively.
Diagnostic tools or outcome measures
The principal biomechanical “tool” still used by most chiropractors is palpation. As such, there has been a continued investigation into factors that change what is felt during palpation. Humans are relatively good at discriminating different magnitudes of stiffness for purely “elastic” materials.  However, the human spine responds as a viscoelastic system, in which the speed of force application changes the apparent stiffness. Nicholson et al  have shown that the relatively poor ability of clinicians to accurately estimate spine stiffness magnitudes is likely due to a 50% poorer ability to discriminate viscous components of viscoelastic systems. Latimer et al  found that therapists used different forces to discern spine stiffness and, hence, had different internal perceptual scales. By training therapists to use a calibrated stiffness instrument, discrimination of PA stiffness in the spine can be done with relatively high interexaminer reliability.  Furthermore, objective instruments have been developed that can reliably measure PA spine stiffness.  Perhaps, the most important aspect of using palpation to detect subluxations (ie, a “manipulable lesion”) is standardization of training.  When examiners are trained in a standardized fashion, they are able to obtain relatively high interexaminer reliability (? = 0.68) for detecting cervical fixations.
Stiffness of the spine is influenced by many factors. If the ribcage is constrained, then the stiffness measured at T12-L4 can be significantly increased.  Change in orientation of an applied load to the spinous process can have small yet significant changes in objectively measured stiffness.  Furthermore, because the spine is a viscoelastic system, there will be a preconditioning effect when applying loads, such that after preconditioning the spine with standard mobilization SM, there will be no measurable change in stiffness.  There has also begun to be a growing appreciation for the natural (and normal) variability in spine stiffness as assessed by standard ROM tests during a physical examination. Christensen and Nilsson  found that in asymptomatic volunteers during a 3-week period, there was an intrinsic variability in ROM of the cervical spine of ± 20°, ± 14°, and ± 12° for flexion/extension, lateral bending, and rotation, respectively. In contrast, repositioning the head to the neutral position, which is related to proprioception, is done with relatively high fidelity over the same period.  Asymptomatic volunteers were able to reachieve the neutral zero position of their heads with a mean difference of 2.7°, 1.0°, and 0.7° for the sagittal, horizontal, and frontal planes, respectively.
Using a case study approach, Lehman and McGill  observed that a single HVLA SM session in the lumbar spine caused notable changes in biomechanical factors associated with a complex task (ie, a golf swing in an experienced golfer who had chronic low back pain). In addition to changes in vertebral kinematics, they observed decreased electromyographic (EMG) responses of the associated lumbar muscles. In a subsequent study, Lehman and McGill  found that lumbar HVLA SM in patients with low back pain resulted in variable changes in lumbar ROM and associated muscle EMG. The largest changes were associated with patients with the greatest reported pain. In a review of the available literature. Lehman  reported that, currently, the best way to discriminate between normal and low back patient groups was using biomechanical tests that assessed “higher-order kinematics during complex movement tasks.” Simpler end ROM tests had poor predictive ability.
Another commonly performed clinical test is measuring leg lengths, especially in the prone position. Using a special designed table to minimize friction and allow independent loading of each leg, Jansen and Cooperstein  determined that the prone leg length test was reliable for detecting non–weight-bearing asymmetry in leg lengths. Nguyen et al  found that there was reasonable concordance (? = 0.6) in determining whether a short leg was present using the Activator protocol. Cooperstein et al  found that it was possible to detect a leg length difference of 1.9 mm but recommended that only differences of greater than 3.7 mm should have confidence associated with them.
One of the signs of maturity of any field is the ability to produce predictive models. In spine biomechanics, most models are computationally based and either use finite element approaches  or optimization with minimization of an objective function. Analytical approaches have also been performed, which include a linear elastic model of a lumbar motion segment.  This model successfully predicted loads born by various ligaments under physiologic loads. Solinger  created a model that predicted the dynamic response of L2-L3 to impulsive loads on the order of those used in VHVLA SM. Using a lumped parameter approach, Keller and Colloca  created an analytical model that predicted the frequency dependent response of the human lumbar spine to PA forces applied to the spinous processes, as is done during low velocity and low amplitude (ie, mobilization), HVLA, and VHVLA (ie, Activator). An alternative approach was adopted by Dulhunty  who modeled force transmission in the cervical spine to predict whether parallel forces or concurrent forces are the optimization function. A relatively new approach in spine modeling, especially in the lumbar spine, is to incorporate what are called “follower loads” for muscles. The issue is that the ex vivo (cadaveric) intact lumbar spine will buckle under compressive loads of ~100 N, whereas in vivo, the lumbar spine easily supports compressive loads of ~1000 N (ie, 10 times greater). Patwardhan et al  found that by modeling muscle activation so that their loads followed the tangent of the lumbar lordosis, their model would approximate the in vivo condition.
A couple of new comprehensive models have been advanced to explain how the spine becomes subluxated in the first place and how SM can restore it to “normal.” Triano [201, 202] has advocated a mechanical model based on the concept of intersegmental buckling, which was based on original observations by Wilder et al [203, 204] and fluoroscopic recordings of a buckling event in a weightlifter by Cholewicki et al  and Cholewicki and McGill.  Essentially, this model proposes that there is a balance point between each pair of vertebrae that under certain loading conditions can suddenly shift, which then results in increased tissue strain of associated soft tissues (eg, facet joint capsule). The increased tissue strain can result in small tears and associated biologic inflammatory response. Evans et al  have proposed an optimization model where the spine system is biased toward minimizing the mechanical energy associated with loading the spine. Their model is described for the case of linear elasticity, although they claim it is also apropos of nonlinear elasticity. As with any theory (or model), the value of these new theories is really found in their predictive ability and how well their predictions are validated by experimental data. So far, neither of these theories has been tested to any degree.
Passive devices have been used for many decades to treat patients with back disorders. Recently, a simple distraction device, Rola Stretcher (Unique Relief, Inc, Davenport, Iowa), designed to be used at home without supervision, was tested to determine whether it showed any lengthening of the spine subsequent to its use. Devocht et al  tested 12 asymptomatic adults and found a significant increase in sitting height after 10 minutes of lying supine on the device. They concluded that it at least temporarily lengthened the spine, presumably by increasing the intervertebral disk height.
In addition to the activator adjusting tool, which has had increasing amounts of scientific study, [160-163] the PulStar computer-assisted, differential compliance spinal instrument has been developed, and a few studies on it have appeared. [209, 210] This latter device also applies an impulse load (up to ~150 N), although the duration of the impulse has not been characterized in articles available in the indexed peer-reviewed literature. The device also incorporates a sensor to measure the compliance of the material that it loads, and hence, the compliance of the paraspinal region can be assessed as well as loaded with the same device. A case study has reported that the instrument was used to treat the spines of infants having colic. 
Determine (quantify) the biomechanical basis of the subluxation.
Determine the parameters that dictate whether a given vertebra should be manipulated.
Determine the parameters that will guide the optimal approach to administering the manipulation.
Determine the effects of manipulation on tissues of the spine.
Which ligaments (including facet joint capsule) sustain the largest strains due to SM
The influence of the vector direction of a given type of SM on ligament strains
Measure the effects of SM on change in tissue characteristics (eg, ligament modulus of elasticity) and cellular response to SM.
Quantify the biomechanical safety of SM in fracture, disk lesions, ligament strains, muscle, and tendon strains.
Develop comprehensive models of the spine that predict how it responds to physiologic and SM loads.
Determine the biomechanical parameters of SM that dominate the neurophysiologic beneficial effects of SM.
Somatic Nervous System
Knowledge of and research directions for understanding the effects of chiropractic spinal adjusting on the somatic nervous system needs, as its basis, an understanding of neurophysiology as it relates to structure and function of the vertebral column. Thus, 2 areas are presented in this portion of the white paper. The first area beginning immediately below represents a substantial portion of our knowledge base for understanding the neurophysiologic properties of paraspinal tissues. The second area beginning with the section on Effects of SMs on Muscle and Muscle Spindles reviews how neural elements of the vertebral column and their organization are affected by SM. Information is included that predates the 1997 white paper when it was not included in that article.
A. Experimental Models to Study Somatic Inputs from the Paraspinal Tissues
Since publication of the original white paper, 2 experimental animal models have been developed that facilitate study of the relationship between spinal biomechanics and neurophysiology in general and of SM specifically: a cervical spine model developed by Bolton and Holland  and a lumbar spine model developed by Pickar.  Additional animal models also relevant to chiropractic spinal adjusting are presented in the section on Animal Models later in this paper. The experimental preparations enable application of controlled mechanical loads to individual vertebra and, at the same time, provide access to the dorsal roots for recording neural activity from paraspinal tissues affected by the mechanical load. The discharge properties of primary afferents with receptive fields in paraspinal tissues and the effects of these sensory inputs on somatomotor, somatovisceral, and central neural processing can be determined. The preparations use a servo-driven motor to control the displacement of or force applied to the spinous process.
Recently, a large animal model (goat) has been used to determine how strains in the facet capsule affect neural input from the capsule.  This model needs additional work to determine whether the capsule is sufficiently preloaded to enable accurate determination of strain and to confirm that identified neurons can be distinguished in the multiunit recordings.
The preparations described in this section provide the opportunity to conduct neurophysiologic studies not possible in humans. With information obtained from these animal models, hypotheses can be formulated and then tested noninvasively in humans.
Group I and II afferents are primary sensory neurons that convey information to the central nervous system from muscle spindles, Golgi tendon organs, and other low threshold mechanoreceptors such as Ruffini endings and Pacinian corpuscles. These afferents conduct action potentials rapidly (>35 m/s) due to their large diameters and heavy myelination.
The structure and function of muscle spindles in the vertebral column have some unique aspects compared with those in the appendicular skeleton. Studies in animal models have described muscle spindles in the hind limb as single receptors located both deep in the muscle belly and close to the musculotendinous junction. [214-217] Spindle densities range from 5 to 45 spindles per gram of hindlimb muscle weight.  In the cervical spine of the human  and cat, [220, 221] however, muscle spindles are rarely seen as single entities, and their densities are greater than in peripheral musculature. In the cat, Richmond and Abrahams  describe cervical spindle complexes wherein 2 to 6 spindles are in close contact with each other or share capsules and/or intrafusal fibers. Spindle density can be 2 to 8 times higher (47-107 spindles per gram) in superficial cervical muscles  and 10 to 25 times higher (137-460 spindles per gram) in deep cervical muscles  than in hindlimb muscles. These differences in spindle densities between axial neck muscles and appendicular muscles appears similar in the humans. 
In the lumbar spine of the cat, Carlson  identified muscle spindles in the longissimus, iliocostalis, sacrocaudalis, intertransversarii, multifidus, and interspinalis muscles, but quantification and morphological description of the spindles were not performed. Similarly, muscle spindles have been identified in the medial, intermediate, and lateral portions of the lumbar erector spinae in the human fetus.  Carlson  also noted that spindle density appeared higher in the central compared with peripheral portions of the longissimus. The high spindle density in the cervical and lumbar muscles is consistent with the high percentage of slow twitch fibers found in muscles of these 2 regions. [220, 223]
The reflex organization of sensory input from paraspinal muscles spindles also has some unique aspects compared with that of the appendicular skeleton. A well-recognized concept related to the cat hindlimb is that the monosynaptic stretch reflex is elicited by excitation of muscle spindles. Afferents from each muscle spindle synapse upon a-motoneurons to that same muscle (homonymous a-motoneurons). [225-227] This stretch reflex arc uses a single excitatory synapse to homonymous a-motoneurons. [226, 228] The afferent arm of the reflex is comprised primarily of group Ia and possibly group II afferents. [227, 229] Each group Ia afferent from a given hindlimb muscle makes functional, monosynaptic connections with 50% to 100% of the homonymous a-motoneurons. [230, 231] Thus, stimulation of a group Ia afferent from a specific hindlimb muscle evokes a monosynaptic excitatory postsynaptic potential in all a-motoneurons to the same muscle. [232, 233] In contrast, in the cervical spine, the monosynaptic reflex connections to homonymous a-motoneurons are weaker. Excitatory postsynaptic potentials are smaller in amplitude, and group Ia afferents make functional connections with only 10% of the homonymous a-motoneurons. [234, 235] This probably contributes to the absence or weakness of monosynaptic reflexes in cervical muscle.  In the lumbar spine of the cat, stretch reflexes can be elicited from the longissimus muscle but not from the iliocostalis muscle. Conduction delays suggest that the reflex arc is not monosynaptic  unlike that in the hindlimb. [226, 228] The presence of monosynaptic stretch reflexes from the deeper lumbar muscles has not been determined. In humans, indirect evidence for the presence of muscle spindles and muscle spindle reflexes in lumbar paraspinal muscles was obtained by measuring evoked cerebral potentials in response to vibration of the lumbar paraspinal muscles,  which relatively selectively stimulates muscle spindles. 
Muscle spindles, along with Golgi tendon organs, comprise a proprioceptive feedback system, which contributes to the sense of movement and position. [240-242] Abnormal sensory input from muscle spindles elicits limb lengthening illusions. [240, 243] When a vibrating mechanical stimulus (100 Hz) is applied to the Achilles tendon of a person standing erect with eyes closed, primary endings in the muscle spindle are excited. Because they monitor change in muscle length, the increased neural discharge signals to the central nervous system that the calf muscles are stretched or lengthened more than they actually are. Spindles increase their static firing rate by ~4 to 5 Hz per millimeter of muscle lengthening.  Because calf muscles normally lengthen as the body leans forward, the proprioceptive feedback error arising from the vibratory stimulus elicits a postural compensation in the form of backward sway. This movement compensates for the illusory forward flexion at the ankle. Recently, Wise et al  showed that spindles in muscles surrounding the elbow are sufficiently sensitive to signal 0.05° to 0.15° changes in elbow rotation. Thus, it seems reasonable to suppose that paravertebral muscle spindles can signal extremely small positional changes or movement of the vertebra to which their parent muscle is attached and, thus, contribute to control of intervertebral motions that might minimize or prevent noxious spinal loading.
Recent findings in humans suggest that proprioceptive input from paravertebral muscle spindles is important for normal reflex activity and repositioning of the lumbar spine. For example, tapping the erector spinae muscles normally elicits short latency paravertebral EMG activity. However, vibration of the lumbar paravertebral muscles, which increases background spindle discharge, inhibits this reflex response.  Additional evidence indicates that proprioceptive input from spindles in the lumbar paravertebral muscles is necessary to accurately position the pelvis and lumbosacral spine. Although healthy individuals can accurately reposition their lumbosacral spine, their repositioning ability is impaired when muscle spindle discharge is increased by applying vibration to the lumbar paravertebral muscles. [246, 247] The correct position is consistently undershot because of the misperception of vertebral position. Interestingly, lumbosacral repositioning ability is impaired in individuals with a history of low back pain, but is improved in the presence of vibration, unlike normal individuals. 
Proprioceptive input can alter muscle force directly via its effect on a-motoneuron excitability and indirectly via its effect on the excitability of segmental and suprasegmental interneurons. Even small changes in paraspinal muscle forces are thought to have a large impact on a motion segment’s biomechanical behavior and stability.  For example, in vitro experiments accompanied by a modeling approach, which incorporated graded activity of 1 lumbar paraspinal muscle, showed an increase in vertebral stabilization as determined by decreases in the intersegmental neutral zone and ROM. Similarly, very small increases in lumbar paraspinal muscle activity at L2-L4 (1-3% of maximal voluntary contraction) were sufficient to restore segmental stability to the lumbar spine even under strenuous loading conditions.  More complex modeling that incorporates force vectors from 5 paraspinal muscles suggests that neuromuscular  mechanisms controlling multifidus muscle activity alone could functionally impact a lumbar motion segment especially during flexion-extension and axial rotation.
A recent study suggests the presence of a previously unrecognized phenomenon in the lumbar multifidus and longissimus muscles that could affect proprioceptive mechanisms controlling paraspinal muscle function.  Changes in intersegmental positions in the lumbar spine that elongated the paraspinal muscles for 10 seconds desensitized paraspinal muscle spindles to subsequent vertebral movement when compared with intersegmental positions that shortened the paraspinal muscles. The findings suggested that either voluntary static postures or involuntary intervertebral positions, which are maintained for short durations, could elicit proprioceptive feedback errors and alter paraspinal muscle force. The spine may be particularly susceptible to this phenomenon because intersegmental positions are not under voluntary control, and a vertebra’s spatial position is not uniquely determined at low loads. 
Group III and IV afferents from deep tissue (labeled as A-d and C-fibers, respectively, when from skin) are primary sensory neurons with mechanically, chemically, or thermally sensitive receptive endings. Some endings are sensitive to only a single modality; others are polymodal. Group III and IV mechanoreceptive endings can have high or low thresholds to mechanical stimuli. Those group III and IV endings that respond in a graded fashion to any stimulus that threatens or actually inflicts injury are called “nociceptors.” Group III and IV afferents conduct their action potentials slowly (=30 m/s) because of their small diameters and light myelination (group III) or lack of myelination (group IV).
Deep tissues of the low back are innervated by afferent endings responsive to both mechanical and chemical stimuli. [253-259] For example, Cavanaugh et al  recorded multiunit activity from group III and IV afferents from the medial branch of the dorsal rami from deep connective tissue after removing lower back muscles in the rat. Gentle probing of the facet capsule, as well as forceful pulling on the supraspinous ligament, elicited a slowly adapting discharge from these afferent nerves. In a systematic study of 57 unmyelinated afferents from the tail and lumbar region of the rat, Bove and Light  found mechanonociceptive endings in muscle bellies, tendon, subcutaneous tissue, and neurovascular bundles. Up to a third of the afferents had receptive endings in more than 1 tissue. No receptive fields were found in the facet joint capsule. Pickar and McLain  recorded single-unit activity from group III and group IV afferents in the intact lumbar spine of cats during movement of the L5-6 facet joint. Most afferents, including 7 with receptive fields in or near the facet joint capsule, responded in a graded fashion to the direction of a nonnoxious load applied to the joint. Yamashita et al  found that only 20% of group III afferents in and around the lumbar facet joint had high mechanical thresholds (>8.5 g), as determined with von Frey–like hairs. This latter finding contrasts with afferents studied in the cervical spine where almost all group III afferents studied had high mechanical thresholds.  In addition, Bolton and Holland  found silent afferents innervating the cervical facet joints, which were only activated by firm, potentially noxious prodding of their receptive fields.
Most unmyelinated mechanonociceptive afferents are also sensitive to chemical stimulation by capsaicin, but only 50% were sensitive to the inflammatory agent bradykinin.  Group III and IV receptive endings in and around the lumbar facet joint can be both activated and sensitized by chemical stimuli. Substance P increases their resting discharge by 80% and decreases their von Frey thresholds by -30%.  Similarly, carrageenan-induced inflammation increases the resting discharge of group III, group IV, and some group II afferents innervating the lumbar muscles and facet joints and sensitizes their receptive endings to mechanical stimuli.  The inflammation also activates previously silent group III and IV afferents.  In the cervical spine, group III afferents with a resting discharge were insensitive to the inflammatory mediator bradykinin,  but previously silent small-diameter afferents were activated by bradykinin.  These neural responses to inflammation likely underlie the findings that mustard oil induced inflammation elicits muscle activity in the neck. Mustard oil intensely activates high-threshold C-fibers (group IV afferents).  When very small volumes (20 µL) were injected into deep cervical paraspinal tissues, EMG activity was increased in a wide variety of upper cervical muscles including digastric, masseter, trapezius, and rectus capitis posterior.  Because the volume was small and its spread to the neighboring tissues was limited, the effects were likely mediated by a reflex. The large number of muscles affected by inflammation of cervical paraspinal muscles may relate to the hyperconvergence, described by Gillette et al  (see next paragraph), and to the communication between segmental paraspinal tissues via intersegmental connections within the spinal cord, reported by the laboratory of Pickar. 
Dorsal horn neurons in the spinal cord with receptive fields in the lumbar paraspinal tissues, including paraspinal muscles and facet joints, receive more convergent input from group III and IV afferents than is true for dorsal horn neurons with receptive fields in the limbs.  In these electrophysiologic studies, Gillette et al  found that wide dynamic range and nociceptive specific neurons in the superficial dorsal horn of the L4-5 spinal segments shared receptive fields with deep and superficial tissues of the lumbar spine, the hip, and proximal leg. This type of input was termed hyperconvergent. Axonal tracing studies revealed that small diameter primary afferents from multifidus muscle and facet joints produce substantial bilateral labeling in laminae I, II, and III, as well as in the deeper laminae V-VIII and X.  Many of these laminae are involved in nociceptive processing and also project to autonomic centers.
Adhesions, fixations, or discal herniation may produce an ectopic source of neural activity. Bove et al  inflamed the axons of mechanically sensitive group II, III, and IV afferents that innervate both superficial and deep structures. The inflammation led to increased spontaneous activity and/or increased mechanical sensitivity of only the group III and IV axons innervating deeper structures.
Increasing evidence shows that the mechanical and chemical consequences of a herniated disk can affect neural tissue within the IVF. Dorsal roots and dorsal root ganglia (DRG) are more susceptible to the effects of mechanical compression than are axons of peripheral nerves because impaired or altered function is produced at substantially lower pressures. [267, 268] Applying as little as 10 mm Hg of pressure to the dorsal root reduces by 20% to 30% the nutritional transport to peripheral axons.  Recently, a mean pressure of 53 mm Hg (range, 7-256 mm Hg) was measured between a herniated disk and the nerve root in 34 humans undergoing surgery for lumbar disk herniation.  Song et al  inserted small pins into the IVF to model a space-reducing lesion in an animal model. Although pressures in the IVF were not measured, this lesion produced mechanical hyperalgesia in the hindlimb and increased the excitability of dorsal root ganglion cells.
The application of nucleus pulposus to a lumbar nerve root increases spontaneous nerve activity and increases the mechanical sensitivity of dorsal root ganglion cells.  In addition, nucleus pulposus applied to a lumbar nerve root produces mechanical hyperalgesia,  causes swelling in and decreases blood flow to the DRG, and decreases blood flow to the lower leg.  Moderate doses of phospholipase A2, an inflammatory mediator associated with disk herniation, also increases the mechanical sensitivity of dorsal roots, produces long-lasting discharges, and increases the discharge of previously silent dorsal root ganglion cells. [41, 274] As mentioned in the previous section, extruded nucleus pulposus contains high levels of TNF-a, and 2 studies have shown that the inhibition of TNF-a by a monoclonal antibody (Remicade) is successful in alleviating sciatica. [74, 75] It should be noted that several case studies [201, 275, 276] and randomized clinical studies [277, 278] show that patients with herniated intervertebral disk, who received SM, gained clinical improvement.
Spinal manipulation induces somatomotor changes, that is, changes in muscle activity, apparently because of sensory input from the somatic nervous system. In asymptomatic patients, Herzog’s group [279, 280] showed that PA spinal manipulative treatments applied to the cervical, thoracic, lumbar, and sacroiliac regions increased paraspinal EMG activity in a pattern related to the region of the spine that was manipulated. The EMG response latencies occur within 50 to 200 milliseconds after initiation of the manipulative thrust. Similarly, SM using an Activator-adjusting instrument applied to a transverse process elicited paraspinal EMG activity at the same segmental level but within 2 to 3 milliseconds.  This is surprisingly fast for a reflex response. Colloca and Keller  confirmed these latter findings in symptomatic patients with low back pain and, in addition, reported that the increased EMG activity, while beginning within 2 to 3 milliseconds of the manipulation, reached its peak within 50 to 100 milliseconds. Paraspinal EMG responses were greatest in magnitude when the manipulation was delivered close to the electrode site, and interestingly, the more chronic the low back pain, the less the EMG response. The EMG electrodes were not placed relative to any physical finding in the low back such as palpable muscle tension, as perceived by the practitioner or tissue tenderness as experienced by the patient.
Spinal manipulation’s effect on paraspinal muscle activity is not exclusively excitatory. In 1 symptomatic patient with spontaneous muscle activity in the thoracic spine, Herzog’s group  observed reduced paraspinal EMG activity within 1 second after a thoracic SM. In a case series study, DeVocht et al  collected surface EMG activity from 16 participants in 2 chiropractic offices. Electrodes were placed over 2 sites exhibiting paraspinal muscle tension determined by manual palpation. Spinal manipulation was administered to 8 participants using Activator protocol. The other 8 were treated using Diversified protocol. EMG activity was decreased after treatment by both methods by at least 25% at 24 of the 31 EMG recording sites.
The effects of SM on paraspinal EMG activity may also be associated with increases in muscle strength. Suter et al  studied symptomatic patients with sacroiliac joint dysfunction, anterior knee pain, and evidence of motor inhibition to knee extensor muscles. A side posture SM applied to the sacroiliac joint significantly decreased the inhibition of the knee extensors on the side of the body to which the manipulation was applied. Similarly, Keller and Colloca  found that erector spinae isometric strength (assessed using EMG) was increased after spinal compared with sham manipulation.
A series of studies have addressed how SM affects central processing of somatomotor information. Spinal manipulation can increase the excitability of motor pathways in the central nervous system and depress the inflow of sensory information from muscle spindles to these motor pathways. This may, in part, account for the disparate clinical findings described above. In asymptomatic patients, Dishman et al  showed that SM increased central motor excitability. EMG activity from gastrocnemius muscle, evoked by direct activation of descending corticospinal tracts using transcranial magnetic stimulation, was larger after lumbar SM compared with simply positioning the patient but not applying the manipulation. However, SM can also depress the H-reflex. Manipulation applied to the sacroiliac joint in a PA direction decreased the magnitude of the tibial nerve H-reflex for up to 15 minutes in asymptomatic humans.  Similarly, side-posture lumbar manipulation of L5-S1 joint inhibited the H-reflex from the tibial nerve.  Mobilization alone but not massage also inhibited the tibial nerve H-reflex, but the effect of manipulation tended to be greater. [288, 289] After manipulation alone, the inhibition lasted for approximately 20 seconds but lasted up to 1 minute when the SM was preceded by spinal mobilization. Similarly, SM delivered to the cervical region depressed the median nerve H-reflex.  The magnitude of the response from the lumbar manipulation was greater than the response from the cervical manipulation, suggesting that central processing of sensory inputs from a SM is different in the neck and the low back.  The depression of the H-reflex does not appear to be a global response. Instead, it appears specific to the region of the spine manipulated because cervical manipulation did not affect the tibial nerve H-reflex.  Patient positioning, which flexes the lumbar spine before the manipulation, may augment the inhibition of tibial nerve the H-reflex. 
A possible mechanism contributing to SM’s inhibitory effects on the H-reflex and on spontaneous paraspinal EMG activity is suggested by recent experiments. Sensory input from tissues of the facet joint elicited by SM might reflexively decrease paraspinal muscle activity. Indahl et al  elicited reflex longissimus and multifidus EMG activity by electrically stimulating the intervertebral disk in a porcine preparation. Stretching the facet joint by injecting 1 mL of physiologic saline abolished the EMG activity.
Haldeman’s group [238, 294] has shown that SM can also affect higher centers in the brain. Using magnetic stimulation, Zhu et al  stimulated lumbar paraspinal muscles and recorded the evoked cerebral potentials. Stimulation of paraspinal muscle spindles using vibration reduced the magnitude of the cerebral potentials. Similarly, muscle spasm in human patients reduced the magnitude of the paraspinal muscle evoked cerebral potentials. Spinal manipulation reversed these effects, reducing muscle spasm and restoring the magnitude of the evoked cerebral potentials. 
There is reason to believe that stretching the facet joint capsule and surrounding tissues likely occurs during SM, although this has received little study.  Furthermore, there may be reason to believe that the mechanically sensitive primary afferents could be stimulated beyond the short duration of an SM. Using MRI scans in human subjects, Cramer et al [10, 11] showed that a side-posture SM accompanied by cavitation gaps the facet joints. The synovial space of the lumbar facet joints increased in width an average of 2.2 mm in subjects who were positioned in side posture and received a side posture spinal adjustment. By comparison, the joint space widened by only 1.5 mm (a difference of 0.7 mm) in subjects who were positioned in side posture but did not receive a manipulation. The MRI scan was performed immediately after manipulation and lasted 20 minutes. Although not studied directly, it seems likely, based upon data from the laboratory of Khalsa,  that joint separations of these magnitudes are sufficient to load the facet joint tissues. If so, this raises the possibility that tissues surrounding the facet joint could be stretched for periods longer than the duration of the manipulation itself. Sensory input from tissues surrounding the facet joint that is graded with direction of facet movement  could elicit reflex muscle responses similar to those measured by Indahl et al. 
Direct evidence from 1 of the experimental models described at the outset of this section  shows that the impulse load of an SM activates a variety of low-threshold mechanoreceptors in paraspinal muscles and that abrupt changes occur in the discharge from their parent afferent neurons [251, 295] as the speed of delivery approaches that used in clinical practice. [296-299] Pickar and Wheeler  recorded afferent activity from muscle spindle and Golgi tendon organ afferents having receptive fields in the lumbar multifidus and longissimus muscles while applying a spinal manipulative-like load to a lumbar vertebra. Muscle spindle afferents from lumbar multifidus and longissimus muscles were stimulated more by the impulse of an SM than by the load preparatory to the impulse (200% compared with 30%). Another type of low-threshold mechanoreceptor, a presumed Pacinian corpuscle, uniquely responded to the impulse of a manipulative-like load, that is, it did not respond to loads with a slower force-time profile. When an SM’s duration was varied between 25 and 800 milliseconds, durations shorter than 400 milliseconds produced abrupt increases in discharge rates from 6 low-threshold mechanoreceptive afferents innervating the lumbar multifidus and longissimus muscles.  An increase in loading magnitude did not appear to systematically affect the discharge from these 6 low-threshold mechanoreceptors. Interestingly, Gillette et al [263, 265] showed that both weak and strong mechanical stimuli applied to paraspinal tissues can suppress spinal cord neurons that receive noxious input from the low back. In an anesthetized human patient undergoing an L4-L5 laminectomy, SM of the lumbosacral region evoked multiunit activity from the intact S1 nerve root.  This neural discharge measured in a clinical setting may be analogous, at least in part, to the discharge of low-threshold mechanoreceptors measured in an animal model.
Numerous studies suggest that SM alters central processing of noxious stimuli because pain tolerance or pain threshold levels can increase after manipulation. In patients with low back pain, Glover et al  examined areas of lumbar skin that were painful to a pinprick. Fifteen minutes after SM of the lumbar region, the size of the area from which the pinpricks evoked pain was reduced, compared with the control group receiving detuned short wave therapy. Terrett and Vernon  quantified the reduction in pain sensitivity after SM using graded, electrical stimulation of cutaneous paraspinal tissues. A blinded observer assessed the minimal current necessary to evoke pain (pain threshold) and the maximal tolerable current that evoked pain (pain tolerance) in subjects with tender regions of the thoracic spine. Spinal manipulation significantly increased pain tolerance levels 1.5-fold within 30 seconds. Over the next 9.5 minutes, tolerance levels progressively increased up to 2.4-fold.
In a case study, Vernon  assessed pressure/pain thresholds before and after SM using a handheld pressure algometer. The threshold measurement indicated the amount of pressure at which the perception of pressure changed to the perception of pain. The algometer was applied to 6 tender points in the neck region. The participant identified his own specific tender points. Spinal manipulation increased pressure-pain thresholds and decreased pain sensitivity by approximately 45% on average. In an effort to extend the findings in this case study, Cote et al  focused on chronic mechanical low back pain. A pressure algometer was applied to 3 sites in the lumbar region. The sites were standardized myofascial trigger points associated with low back pain. Unlike Vernon’s earlier case study, these trigger points were not necessarily clinically relevant, that is, they were not identified as tender by participants, nor were they necessarily the most sensitive points for each individual. Unlike Vernon’s case study, no changes in pressure-pain thresholds were observed.
“Nearly all proposed theories to explain the effects and mechanisms of action of SM have failed to withstand intense scientific scrutiny” (from the original white paper, Brennan et al ).
This statement from the original white paper does not adequately express the state of chiropractic science.
It seems more accurate to say that nearly all the proposed theories to explain the effects and mechanisms of action of SM have not been fully tested.
Spinal manipulation is a biomechanical input generally delivered at high velocity. The question of how a short-lasting biomechanical input can presumably have long-lasting changes on a person’s health needs answering. Research should seek to determine if SM produces long-term effects on biomechanics and/or neurophysiology.
Determine the discharge characteristics (ie, the pattern or frequency of action potentials) of primary sensory neurons innervating the vertebral column in response to high-velocity loading.
Determine how these patterns of activity affect the signaling properties of neurons in the central nervous system, for example, do they produce long-lasting changes.
Determine if SM produces long-lasting changes in spinal biomechanics, which would presumably produce long-lasting changes in sensory input.
Determine if SM produces long-lasting changes in neuromuscular control of paraspinal muscles, possibly comparing the use of fine wire electrodes with surface EMG or scanning EMG.
Determine if paraspinal tissues have any unique physiology, compared with appendicular tissues, by comparing, for example, reflex changes initiated from sensory receptors in appendicular tissues with reflex responses initiated from sensory receptors in axial tissues.
Identify objective changes in the vertebral column that lead one to think that SM is needed.
Use new technologies to determine changes in intersegmental stiffness.
Determine if hyperalgesia is associated with the manipulable lesion.
Determine whether the manipulable lesion is inflamed and, conversely, if inflammation of spinal and paraspinal tissues can cause the manipulable lesion.
General Recommendations. A seed recommendation in the 1997 white paper (Brennan et al ) was to “concentrate basic science research at one lead chiropractic college and provide sufficient support personnel to conduct needed studies.” Acting on this recommendation in some way is sorely needed. A critical mass of chiropractic-oriented scientists able to easily interact and develop new ideas is critically important to drive chiropractic science. Basic and clinically-oriented basic scientists are scattered almost individually at chiropractic colleges and are isolated. This is in sharp contrast to the traditional university setting where a wide variety of fields are represented and allow for not only interdisciplinary collaborations but provide opportunities for informed discussion leading to inspiration and the writing and submission of research grants. Consideration should be given to the relative merits of developing basic science infrastructure at chiropractic colleges either in contrast to or in addition to establishing a small cadre of chiropractic scientists within a traditional university setting containing established infrastructure in terms of space, equipment and collaborative potential with established scientists from multiple disciplines.
The remarkable scientific progress in both industry and medicine over the last 100 years has been responsible, in large part, to a transition from observational to experimental research. [305, 306] Human clinical studies contributed greatly to this progress, but animal studies have permitted investigators to perform experimental interventions that were not possible in human studies, and allowed a wider range of study designs. In addition, statistical power is easier to obtain in animal studies because large numbers of animals can be evaluated at relatively low costs, and animal study groups can be genetically homogeneous. Moreover, research animals are often bred to have genetic predispositions to illnesses that mimic those of humans, such as asthma,  cancer,  diabetes mellitus,  and hypertension.  These important features, and the ability to strictly control potentially confounding influences, make animal research an essential tool for today’s health care researchers. Consequently, the need for animal models in chiropractic research has been acknowledged in each of the major reviews of scientific progress in chiropractic. [1, 311, 312]
Despite this, some still question how we can learn anything about people from studying animals. The basis for this concern lies in the obvious anatomical and physiologic differences between humans and animals. Animal studies are generally used to examine fundamental mechanisms that are common to both humans and nonhuman species. In addition, as noted above, many human diseases can be mimicked in animal models. Consequently, animal research provides information about fundamental mechanisms common to both humans and animals, and often suggests new hypotheses for evaluation in subsequent human studies. The discovery of insulin provides an excellent example. It was also one of the most dramatic events in the history of health care research.  Animal studies showed that the pancreas was a critical organ in the development of diabetes mellitus. Additional work produced an extract of the pancreas that reduced hyperglycemia and glycosuria in animals that had been previously rendered diabetic by removal of the pancreas. After further extensive evaluation with laboratory animals, the purified extract was deemed ready for human tests. In the first human trial, a 14-year-old boy with severe diabetes received an intramuscular injection of the “purified” pancreatic extract but failed to show clinical improvement and developed a sterile abscess at the injection site.  However, on the strength of the previous animal studies, work continued to further purify the pancreatic extract, and additional human studies were conducted. These new human studies, using the purified extract, showed a tremendous clinical improvement in all subjects.  Therefore, animal studies revealed the critical role of insulin in diabetes, provided a source of the hormone for subsequent study, and showed the potential of insulin as a therapeutic agent. All of these events were necessary before human clinical trials of insulin could begin. In 1923, the Nobel Prize for Physiology or Medicine was awarded to Banting and Macleod, recognizing that the discovery of insulin had  “conferred the greatest benefit on mankind.” The current era of chiropractic experimental research began after the first federally funded workshop to examine SM, The Research Status of Spinal Manipulative Therapy, in 1975.  At the conclusion of this historical conference, it was widely acknowledged that little basic or clinical research data were available to evaluate the claims of clinicians using SM.
Immediately after the first Research Agenda Conference in 1996, a white paper was published on the status of basic science research in chiropractic, “Basic Science Research in Chiropractic: The State of the Art and Recommendations for a Research Agenda”.  A contemporary review by Vernon was cited in which 18 animal studies examined spine subluxation (1 historical monograph, 4 abstracts, and 13 articles).  A recent review by Henderson, Animal Models in the Study of Subluxation and Manipulation: 1964 to 2004, presents synopses of 34 animal studies (5 abstracts and 29 articles) published within the past 40 years.  In this review, studies were included if they specifically examined subluxation, the osteopathic lesion (somatic dysfunction), or SM. Studies examining subluxation or somatic dysfunction were grouped under the general term subluxation studies. These studies used animals to model either subluxation (31 studies) or SM (3 studies). The 31 subluxation studies examined either “full-mimic models” that attempted to induce spine fixation in intact animals (14 studies) or “component models” that emulated specific mechanical or chemical components attributed to spine subluxation (17 studies). The 3 SM studies used either manual (1 study) or instrumental interventions (2 studies).
A. Subluxation Mimic Models
Only 1 subluxation mimic model has been introduced since the 1996 Research Agenda Conference. This model, the external link model, combines surgically implanted spinous attachment units and an external link system to produce reversible, mechanical fixation of 3 adjacent lumbar segments (L4, L5, and L6) in the rat.  Cramer et al  used the external link model to examine degenerative changes after spine fixation. They observed stiffness and Z joint changes that developed within weeks after experimental fixation of a spine segment. Both the occurrence (number of involved segments) and severity (0-3 scale, least to most severe) of degenerative changes were recorded. These investigators reported significant differences in Z joint degeneration between fixed segments and nonfixed segments within the same animal. In addition, the occurrence and severity of articular degeneration and osteophyte formation on Z joints in rats with fixated vertebrae was significantly greater than similar degenerative changes, on comparable segments, in never-linked control rats.
This subluxation mimic study provided strong evidence that decreased vertebral motion (vertebral fixation) produced degenerative changes in the Z joint that were greater for longer periods of fixation. Generally, these degenerative changes continued to progress after removal of the fixating links. However, the data also suggested that time thresholds exist, before which removal of the experimental fixation (links) may spontaneously reduce or reverse the fixation-induced degenerative changes. These time thresholds appeared to be earlier for facet surface degeneration (occurring between 1 and 4 weeks of fixation time) and later for osteophytic degeneration (occurring between 4 and 8 weeks of fixation time). In addition, facet degeneration was observed to occur earlier than osteophyte formation. The existence of these time thresholds is intriguing, and may have clinical significance. However, the authors warned that there is no known basis for projecting rat time frames to human subjects. Further work with this model and subsequent human studies are required to expand our understanding of these issues.
There has been a substantial number of interesting studies using subluxation component models since the first Research Agenda Conference review of basic science in chiropractic. The work of the laboratory of Pickar  stands out because these investigators have produced a continuous line of inquiry, modifying their experimental preparation in the cat to produce multiple subluxation component models (see section on Somatic Nervous System). Pickar introduced a unique surgical approach in the anesthetized cat that allows great diversity in spine loading conditions while recordings are obtained from single-unit afferents within the L6 dorsal root. Pickar found that muscle spindles and Golgi tendon organs located in the local paraspinal muscles respond to lumbar spine loading with unique activity patterns that reflect changes in both the magnitude and direction of the load.  Pickar and Kang  showed that holding the lumbar spine in 2 different static positions for as little as 5 seconds altered the sensitivity of paraspinal muscle spindles. This interesting observation suggests that postural positions before spine motion can bias spindle-based information about spine motion and position. Such a “muscle history” effect may have serious clinical consequences if it introduces sufficient proprioceptive error. The disrupted synergy among and between muscle agonists and antagonists could readily predispose joints to injury. In addition, they reported electrophysiologic evidence that the medial branch of the dorsal ramus is a critical part of the pathway for intersegmental reflexes between lumbar spinal and paraspinal tissues.  Finally, these investigators used their experimental preparation to explore somatovisceral mechanisms associated with subluxation.  They observed changes in heart rate, blood pressure, and splenic and renal sympathetic nerve discharge activity in response to dorsoventral loading of the L3 vertebra with and without multifidus muscle inflammation. Increased heart rate, blood pressure, and splenic and renal sympathetic nerve discharge activity was observed in response to multifidus muscle inflammation but not to dorsoventral loading alone. Combined dorsoventral loading of the L3 vertebra with multifidus muscle inflammation produced smaller but more prolonged, increases in the measured outcomes. This interaction suggests somatovisceral responses that are due to competing excitatory influences from paraspinal muscle inflammation and inhibitory influences from mechanical spine loading. This study from Pickar’s laboratory and the extensive work of Sato, Budgell, and their colleagues [321-324] form an important body of work that has begun to clarify segmental and suprasegmental somatovisceral reflex mechanisms mediating the chiropractic clinical experience.
Budgell and his colleagues, working mostly in the laboratory of Sato, have continued the line of somatovisceral research began by Sato and Swenson.  They have examined changes in mean arterial blood pressure, heart rate, and sciatic nerve blood flow in rats receiving mechanical and chemical stimulation to the forelimb, hindlimb, and spine (L4-L5 facet joint and interspinous ligament) both before and after thoracolumbar spine transection and/or sympathectomy. [321, 322] The cardiovascular responses to noxious and innocuous mechanical and chemical stimuli of spinal structures were different then those observed after stimulating appendicular structures. These studies showed that sciatic nerve blood flow was determined by a competition between dominant supraspinal influences and weaker segmental influences.
This raised a provocative question that is highly relevant to chiropractic theory: “Can noxious stimulation of somatic tissues at one site elicit reflex attenuation of blood flow in more distant nerves with resultant sensory and/or motor disturbances, producing aberrant end organ function?” In recognition of the complexity of the autonomic response they commented,  “These results draw attention to the difficulty in predicting the integrated manifestation of somatoautonomic reflexes. A single stimulus may unleash a cacophony of responses that compete for expression in the end organ.” The work of Budgell et al  showed that chemically irritating lumbar interspinous tissues to model the intensity and chronicity of biomechanical back pain increased sciatic nerve blood flow via both segmental and suprasegmental mechanisms. However, it was not clear that this would produce meaningful changes in organ function. Subsequent work by Budgell showed that irritation of either thoracic or lumbar interspinous tissues produces a strong somatovisceral reflex that increases catecholamine secretion by the adrenal gland as well as marked increases in bladder motility. 
Since the previous Research Agenda Conference white paper on basic science in chiropractic,  subluxation component studies by 2 other basic science researchers have contributed to our understanding of possible mechanisms underlying the clinical chiropractic experience. Gillette et al  extended an earlier work in the cat,  in which they had shown dorsal horn “hyperconvergence” of somatosensory input from different deep somatic tissues and from skin in the low back, hip, and proximal leg (see also Somatic Nervous System). That earlier study showed a neurophysiologic basis for the observation that patients can not readily localize low back pain and suggests a mechanistic explanation for the pattern of hip and proximal thigh pain commonly associated with low back injury in humans. In addition, they had shown, for the first time, that noxious mechanical or chemical stimulation of the lumbar dura mater excites dorsal horn neurons. That finding is consistent with recent studies in which the dura mater has become a focus for understanding neck and back pain and should lead us to reexamine the interpretation of the straight leg, nerve tension test. [326, 327] The follow-up study by Gillette et al  showed that activity in these hyperconvergent dorsal horn neurons driven by noxious mechanical stimulation could be suppressed by both weak and strong mechanical stimulation within their receptive fields. This finding provides a possible mechanism for the clinical observation that SM ameliorates low back pain. 
Most recently, Song et al  used their reversible model of IVF encroachment in rats to examine changes in DRG cell activity and the associated functional consequences. These researchers produced IVF encroachment and related inflammation at L4 and L5 and observed increased excitability in related DRG cells as well as hindlimb hyperalgesia, both of which developed rapidly. Removing the IVF encroachment normalized the DRG cell activity and initiated a rapid recovery of the hindlimb hyperalgesia. These findings are consistent with the IVF encroachment theory posited by chiropractors to explain a mechanistic link between subluxation and diverse clinical consequences. [328, 329]
Since the first white paper on chiropractic basic science, all but 1330 of the published SM mimic animal studies have been performed in the laboratory of Pickar.  Pickar’s unique surgical approach in the anesthetized cat allows introduction of a great diversity of spine loads while recordings are obtained from single unit afferents in the L6 dorsal root. In fact, researchers in Pickar’s laboratory were the first to show that muscle spindles and Golgi tendon organs in paraspinal muscles respond to vertebral loads with force-time profiles similar to SM.  Moreover, the unique response of paraspinal proprioceptors to the impulse (thrust) portion of the SM-like load suggests that these receptors may contribute to the therapeutic effects of SM, possibly explaining changes in spontaneous EMG reported in clinical studies. [279, 288] These investigators recently studied the effect of manipulation impulse rate on low-threshold mechanoreceptors in lumbar paraspinal muscles.  They found that different rates of manipulation produced substantial differences in the discharge frequencies of afferents from low-threshold mechanoreceptors in paraspinal muscles. Afferent discharge frequencies increased abruptly as the impulse duration of the manipulation decreased to 100 milliseconds, a duration typically observed during chiropractic SM. [297, 331] In addition, it was noted that muscle spindles and Golgi tendon organs were coactivated at these higher speed manipulations, but not at slower, mobilization-like, manipulations. These findings are consistent with the report of Herzog et al,  in human subjects, that treatment forces of similar magnitude and direction but applied at different rates (ie, >1 second compared with 100-150 milliseconds) produced markedly different EMG responses. Herzog et al  reported that durations of 100 to 150 milliseconds produced “clear, measurable, and repeatable EMG responses in a distinct area that is specific to the treatment administered… the production of the reflex response depends directly on the rate of change in force and deformation during the treatment rather than on the force of stretch magnitude itself.” These findings from both animal and human studies support the widely held chiropractic perspective that high-speed (short-duration) manipulations have a unique influence on the nervous system, as compared with slower, mobilization, manipulations. See Somatic Nervous System for additional details.
Develop animal research capacity in chiropractic.
The profession’s support for animal research related to chiropractic procedures and practice should be provided to chiropractic institutions that show a clear and substantial commitment to perform this work.
In the previous basic science white paper, it was stated with regard to animal studies, “The major weaknesses of these studies are the lack of replication of most of them and the fact that there has been no comprehensive, focused line of investigation that has resulted in a substantial body of information in the peer-reviewed, indexed literature”.  Since that publication, only a small number of basic science researchers have continued to work with animal models of subluxation and/or SM. The last white paper also recommended, “Concentrate basic science research at 1 lead chiropractic college and provide sufficient support personnel to conduct needed studies”.  The profession has made no concerted effort to act on that recommendation. However, a few chiropractic institutions have established animal research facilities and performed basic science research. Collaborations between their basic research faculty on private and federally funded projects has also strengthened these animal research programs. However, these facilities are expensive to maintain and are not fully used at this time. This low use is largely due to the small number of trained basic research scientists with release time from other responsibilities. Animal research facilities at these chiropractic colleges provide the profession a unique opportunity to capitalize on existing assets.
It has been argued that conventional basic science research institutions already have facilities, support personnel, and research faculty that may be enlisted in the exploration of chiropractic issues. Research faculties at these institutions are well positioned to obtain both federal and private funding. In fact, a substantial portion of research funding from the chiropractic community has been awarded to researchers at these institutions. However, diverting chiropractic’s very limited financial resources into such institutions leaves our own research infrastructure undeveloped. “Foundational” chiropractic research directly examines the theoretical tenets of the chiropractic profession. This is almost always basic (mechanism driven) research as opposed to clinical (outcomes driven) research. Basic, foundational research capacity must be purposely nurtured in chiropractic institutions where chiropractic theory is formed and taught. Consequently, it is particularly important that the profession dedicate funds to strengthen and maintain animal research capacity at chiropractic institutions. To maximize the impact of our limited funds, we must capitalize on existing assets. Therefore, to be eligible for supplemental basic research funding from the profession, chiropractic institutions must show a minimum level of accomplishment in 5 critical areas:
Have an Institutional Animal Care and Use Committee that fully complies with and enforces the published guidelines of the National Institutes of Health Office of Laboratory Animal Welfare and the US Department of Agriculture Animal and Plant Health Inspection Service.
Provide animal-research infrastructure in the form of core facilities for animal care (with vivarium and surgical suite) that are in full compliance with Office of Laboratory Animal Welfare and Animal and Plant Health Inspection Service guidelines.
Equip basic research laboratories that will support biochemical, histologic, and physiologic studies.
Provide 1 or more animal care personnel with either a bachelor’s degree in laboratory animal husbandry, American Association for Laboratory Animal Science credentials, or at least 3 years of animal care experience.
Provide 1 or more basic laboratory personnel with either a bachelor’s degree in science or at least 3 years experience as a laboratory technician.
Have 1 or more faculty with external funding, private or federal, for basic science research using animals.
Perform basic, foundational chiropractic research studies.
Have a history of basic research collaborations with other chiropractic colleges and conventional research institutions.
Establish a Foundations Research Chair (1 full-time equivalent) and a Foundations Research Associate position (1/2 full-time equivalent) at 1 or more of the eligible chiropractic institutions to conduct animal research studies that will explore fundamental subluxation and spine manipulation issues.
Every 3 years, identify a Critical Area of Need that targets a foundational chiropractic research topic and provide funding for a 2- to 3-year collaborative study.
Establish a Supplemental Funds Program to support animal research studies that achieve federal funding in the exploration of chiropractic issues. This program would be available to all investigators examining chiropractic issues.
Provide annual Foundations Research Training Stipends for basic science investigators at chiropractic colleges that have not yet attained the above minimal accomplishment criteria. This funding will permit investigators at unsponsored chiropractic institutions to receive on-site training at the sponsored foundational basic research sites.
Establish a Tissue Bank at sponsored animal research sites that have established long-term survival animal models such that harvested tissue specimens from these studies will be available for both a priori and post hoc analysis. Basic science researchers at off-site research institutions may thus obtain specimens for collaborative studies. Consequently, the Tissue Bank will provide opportunities for collaborative research among chiropractic basic science researchers at numerous research institutions and optimize the use of animals in long-term survival studies.
Specific basic science research topics.
Animal research faculty should conduct studies to describe the nature and effects of spine lesions that respond to chiropractic manipulation and should also examine the specific attributes of the various SM methods and their differential effects.
Observations from subluxation mimic and subluxation component models suggest the reality of the subluxation as an experimental phenomenon, but the existence of this spine lesion in the “real world” outside the laboratory has not been established in either animals or humans. Animal studies have been both informative and provocative, but their small number is presently inadequate as an evidence base for the chiropractic profession.
The current body of animal research also suggests that stimulating spinal and paraspinal tissues produces significant somatic and visceral effects that may be uniquely different from those observed after stimulation of appendicular structures. This is encouraging and consistent with chiropractic theory, but many questions remain unanswered. Future animal studies should examine behavioral consequences of spine subluxation and manipulation as well as diverse anatomical and physiologic outcomes. These additional studies may provide a mechanistic explanation for field practitioner observations that both somatic and visceral problems improve with SM. 
Research history strongly supports the value of animal studies as a window into the biologic mechanisms common to both animals and humans. It is clear that the development and use of animal models for chiropractic basic science research will play a central role in achieving recognition and respect for chiropractic by the world community. Moreover, this powerful tool will help to develop chiropractic theory, inform educators, and enhance the effectiveness of chiropractic practice.
The biomechanical spine lesion
Identify and characterize putative biomechanical spine lesions that respond to spinal manipulative therapy (prevalence, causation, and biologic significance: nature, longevity, magnitude and diversity of effects, and natural history).
Evaluate the deterministic role of specific segmental level and multiple involved segments on the biologic effect of these spine lesions.
Evaluate the potential adverse contributory effect of biomechanical spine lesions on specific health conditions (using animal models with genetic predispositions to illnesses that mimic those of humans; such as asthma, cancer, diabetes mellitus, and hypertension).
Determine to what degree homologues of the experimentally induced biomechanical spine lesions (subluxation mimics) exist in unmodified animals.
Identify and characterize the effects of SM (biologic significance: nature, duration, and magnitude of effects).
Evaluate the differential impact of SM application parameters such as rate (impulse duration), impulse frequency (single vs multiple impulse cycles), load vector (direction and magnitude of applied load), displacement vector (direction and magnitude of displacement), site of application (vertebral or paravertebral), and course of treatment (frequency and distribution of treatments over a treatment period).
Evaluate the potential benefits of SM on specific health conditions commonly seen in chiropractic practice (using animal models that mimic specific human illnesses).
Evaluate therapeutic benefits derived from SM to determine whether they are dependent on correcting a specific spinal lesion (eg, subluxation) or may be independent of a preexisting spine lesion.
Global animal research issues
Evaluate the possible deterministic role of bipedal vs quadruped posture on outcomes associated with biomechanical spine lesions and SM.
Perform integrated animal and human studies that will examine specific biologic effects that have been shown in animal studies and will determine their homology in humans.
Immune Function Studies
Chiropractic field practitioners assert that SM produces improvement in several visceral conditions, including some that are thought to involve perturbed immune function. [332-334] Although chiropractic theory suggests that immunologic effects might be expected through neurologically mediated mechanisms,  few studies have addressed this topic. [317, 329, 336] Most of the evidence has been drawn from basic science studies that gave broad mechanistic support but did not specifically examine SM and its effect on immune function.
The central nervous system and immune system share modulator and receptor mechanisms by which the 2 systems communicate. Their interaction maintains both basal and stress-related homeostasis through 2 major pathways: the systemic sympathetic nervous system (SNS) and hypothalamic-pituitary-adrenal (HPA) axis. The homeostatic role of catecholamines released via the SNS is well known as a major regulator of metabolism, heart rate, blood pressure, and core temperature.  Similarly, the HPA axis has immunosuppressive and anti-inflammatory effects via release of adrenal glucocorticoids.  Within the past 2 decades, numerous studies have also shown direct SNS innervation of both primary and secondary lymphoid tissues in all species studied.  In these tissues, varicose sympathetic terminals form plexuses that run adjacent to vascular smooth muscle and also terminate in parenchymal lymphoid tissue well removed from blood vessels. These findings suggest that the SNS has a nonsynaptic noraderenergic, modulatory influence on lymphoid tissues. In addition, immune cells have been shown to express adrenoreceptors, which modulate lymphocyte trafficking, proliferation, receptor expression, and cytokine production. Cytokines are the hormone-like messenger molecules that immune cells use to integrate the functional activity of other immune cells. 
Consequently, the immune system is now thought to be “tuned” by contrasting neural influences; the local release of norepinephrine from nonsynaptic, postganglionic sympathetic terminals and circulating epinephrine secreted by the adrenal medulla. When internal or external influences disturb homeostasis, both the SNS and HPA axis are activated, thereby increasing the peripheral levels of catecholamines and glucocorticoids to restore the steady state of the internal milieu. In the 1930s, Selye  identified this integrated response as the general adaptation syndrome. Centrally, the 2 principal components of the general adaptation syndrome are the release of corticotropin-releasing hormone and the locus ceruleus–NE/autonomic (sympathetic) nervous system. These 2 systems participate in a positive, reverberating feedback loop in which activation of 1 system tends to activate the other as well.  Activation of the locus ceruleus–NE system leads to release of NE from an extraordinarily dense network of neurons throughout the brain. The functional consequence is increased arousal and vigilance, increased epinephrine release from the adrenal medulla, and increased peripheral sympathetic activity resulting in the release of NE from the varicose sympathetic nerve terminals. Norepinephrine and epinephrine bind to β2-adrenoreceptors on antigen-presenting cells and helper T cells, thereby inhibiting the production of TH1, proinflammatory cytokines (eg, interleukin 2, TNF-a, and interferon-?), while stimulating the production of TH2, anti-inflammatory cytokines (eg, interleukin 4, 5, and 10, and transforming growth factor–β). This shift in cytokine profiles from TH1 to TH2 produces a functional immune shift from a cellular response to a humoral response. Catecholamines may also increase regional immune responses by inducing interleukin 1, TNF-a, and interleukin 8 production. Therefore, sympathetic innervation may serve to localize and focus an inflammatory response via neutrophil accumulation and activation of specific humoral responses while systemically inhibiting TH1 cellular immunity responses. This would also protect the organism from the production of proinflammatory cytokines and other products of activated macrophages.
There is also considerable peptidergic influence contributed by small afferents that are confined mainly to the parenchyma of all lymphoid organs (eg, substance P, neurokinin A, and calcitonin gene-related peptide) and neuropeptides that are colocalized in the large dense-cored noradrenergic vesicles of sympathetic varicosities (eg, substance P, somatostatin, and vasoactive intestinal peptide). [338, 342] A close spatial relationship has been reported between peptidergic nerve fibers and mast cells, T cells, and macrophages in lymphoid tissues.  It is noteworthy that mast cells express receptors for substance P that trigger release of histamine and other factors such as leukotrienes.  Both substance P and histamine induce plasma extravasation and vasodilation in local tissues. In addition, substance P stimulates postganglionic sympathetic terminals to release norepinephrine and prostaglandins E2 and I2.  Consequently, in addition to their direct immunomodulatory effects on immune cells, neuropeptides can exert important indirect immunomodulatory effects via modulation of histamine release from mast cells and postganglionic sympathetic terminals in the parenchyma of lymphoid organs.
Finally, cytokines and their receptors have been shown within the central nervous system. Therefore, catecholamines and neuropeptides can regulate immune function, and cytokines can act as neuromodulators. The cytokines TNF-a, interleukin 1, and interleukin 6 have all been shown to modulate the SNS and the HPA axis. [345, 346]
A. Immune Function and Subluxation/Spinal Manipulative Therapy
In the previous basic science white paper,  a review of the literature found relatively few studies that reported immune function effects associated with chiropractic manipulation or spine subluxation. Vernon et al  reported a slight, but statistically significant, increase in β-endorphin levels in asymptomatic males after cervical manipulation, whereas Sanders et al  and Christian et al  found no change in β-endorphin levels in either symptomatic or asymptomatic male study participants after chiropractic manipulation. Christian et al also reported no change in adrenocorticotropic hormone and cortisol levels between sham and treated groups or between pre- and post-treatment in any group. They concluded that changing β-endorphin levels did not mediate the analgesic response attributed to chiropractic manipulation, and that manipulation did not activate the HPA axis.  The work of Brennan et al [350-357] remains the only extended line of investigation into the effect of chiropractic SM and immune function. They reported that a single manipulation in the thoracic or lumbar spine produced a short-term priming of the polymorphonuclear cell response to an in vitro particulate challenge. They observed an enhanced chemiluminescent respiratory burst in both asymptomatic and symptomatic study participants. [350, 352, 357] This enhanced polymorphonuclear cell activity was associated with slight, but statistically significant, rise in plasma substance P. Further investigation suggested that this systemic effect was dependent on both the applied force and vertebral level. [352, 357] Brennan et al also found that patients presenting with neuromusculoskeletal complaints had reduced numbers of circulating natural killer cells; but these cells were not functionally impaired. [353-355] Lastly, SM did not alter the absolute or relative number of circulating immune cells (eg, B cells, T cells, and natural killer cells) in study participants enrolled in a clinical trial of SM for the treatment of chronic low back pain. 
An article published shortly after the 1997 basic science white paper reported a case series on 9 study participants of an initial 18 participants enrolled in the study.  Each study participant received 4 HVLA manipulation treatments to correct hypomobile vertebral segments over a 2-week treatment period. Saliva was collected for cortisol assay 2 weeks before the treatment period, during the 2-week treatment period, and 1 week after treatment. This was a very small study with poor participant compliance (50%). The study author reported no statistically significant effect of chiropractic manipulation on salivary cortisol levels, unless 1 outlier was removed from the data analysis. The very small sample, poor participant compliance, post hoc analysis, and several design weaknesses (no control, unclear attention to time of day for saliva collection, and possible confounding influences due to female estrogen cycle) make these results equivocal.
A subsequent salivary cortisol study by Whelan et al  examined 30 asymptomatic male chiropractic students in a randomized, 3-arm clinical trial (manipulation, sham, and no treatment control) to determine the effect of HVLA cervical manipulation on salivary cortisol secretion. These investigators were interested in determining whether chiropractic manipulation is itself a physiologic stressor capable of producing a significant stress response in asymptomatic subjects. They found no effect of chiropractic manipulation and concluded that, in asymptomatic subjects familiar with chiropractic manipulation, neither the “set-up” sham nor cervical manipulation induce a state of anxiety sufficient to disrupt the homeostatic mechanisms and activate the HPA axis. Unfortunately, the conclusions of this study can not be readily assessed because the authors did not discuss the statistical power of the study and failed to provide data for the reader to determine whether the study group size was adequate for detecting a clinically significant difference in salivary cortisol levels.
Two very recent articles, an animal study  and a human study  have reported interesting immune system responses after SM.
Song et al  produced a rat model of acute IVF inflammation with associated behavioral hyperalgesia, electrophysiologic changes, and neurohistologic pathology by injecting 30 µL of an “inflammatory soup” (bradykinin, serotonin, histamine, and prostaglandin) directly into the L5-L6 IVF of each of 100 experimental group rats. An additional 48 control rats underwent an equivalent surgical preparation but without injection of the inflammatory soup.
These investigators evaluated the therapeutic effects of 2 weeks of SM therapy (SMT) applied to the L4, L5, L6, or L5 and L6 spinous processes using an Activator II instrument (Activator Methods International). Behavioral hyperalgesia was evaluated before and periodically after IVF injection by recording foot withdrawal latencies to thermal stimuli and thresholds to mechanical indentation with von Frey filaments. They also recorded resting membrane potential, action potential current threshold, and repetitive discharge characteristics, in vitro, from neurons in the L5 dorsal root ganglion. Lastly, L5 DRG were taken from rats at different periods, stained with hematoxylin and eosin, and observed via light microscopy for signs of inflammation (eg, increased vascularization and satellitosis).
Song et al  reported that Activator SMT applied to L5, L6, or L5 and L6, but not L4, significantly reduced the severity and duration of both the thermal and mechanical hyperalgesia produced by IVF inflammation. Their electrophysiologic studies showed that inflammation-induced hyperexcitability of dorsal root ganglion neurons was also significantly reduced by the SMT. Finally, the increased vascularization and satellitosis observed in inflamed L5 dorsal root ganglion were significantly reduced 2 to 3 weeks after the Activator SMT. They concluded that Activator SMT can significantly reduce the severity and shorten the duration of pain and hyperalgesia caused by lumbar IVF inflammation. They also noted that the SMT effects were segment-specific.
The very recent article by Teodorczyk-Injeyan et al  reports that SMT in asymptomatic subjects down-regulates production of the inflammatory cytokines TNF-a and interleukin 1β (IL-1β). They also determined that this change in cytokine production was unrelated to serum substance P levels.
These investigators recruited 64 age- and sex-matched asymptomatic subjects who had not received an SM for a minimum of 6 months. Subjects were equally and randomly distributed across 3 study groups: SMT, sham-SMT, and venipuncture control. SMT subjects each received a single bilateral hypothenar (Carver-Bridge type) thrust applied to an upper thoracic (T1-T6) segment that had been previously identified as having a motion restriction. The SMT was judged to be successful if an audible release (cavitation) was heard. By contrast, the sham-SMT subjects received a similar setup and thrust that was directed such that no audible release was heard. Subjects in the venipuncture control group were treated similarly to the SMT and sham-SMT groups, except that no thrust was given. Blood samples were obtained from all subjects before any intervention (or sham intervention) and at 20 minutes and 2 hours after intervention (or sham intervention). All blood samples were coded to blind laboratory investigators to study group assignment.
Whole-blood cultures were subsequently activated with lipolysaccharide for 24 hours, and TNF-a and IL-1β production was determined from culture supernatants by specific immunoassays. Substance P production was determined from intact sera using a competitive immunoassay.
A statistically significant proportion of sham-SMT and venipuncture control subjects showed progressive increases in the synthesis of TNF-a and IL-1β, whereas a comparable proportion of SMT subjects showed a gradual decline in both cytokines. Subjects in all groups had normal baseline cytokine values within 2 hours after the intervention (or sham intervention). In all study groups, serum levels of substance P remained unaltered from baseline values.
As reported in this white paper, basic science studies support chiropractic theory that spinal subluxation and SM impact neurologic function. In addition, the interdependence of nervous, endocrine, and immune systems has been discussed here. These studies suggest mechanisms by which spinal influences may mediate a clinically significant impact on immune function, but few studies have directly examined those mechanisms, and the evidence to date is conflicting. Same chiropractic field practitioners believe that, in addition to nutritional therapy, SM benefits patients having infection and immune system diseases. [329, 349] Consequently, there are both plausible mechanisms to explore and clinical practice driven justification for additional basic science studies in this area.
Develop immune function research capacity in chiropractic.
The profession’s educational and research institutions lack experienced molecular biology researchers who are familiar with the current immune function literature and research methodology.
Fund subcontracted molecular biology support for immune function studies that will be conducted at the sponsored foundational animal research sites (see “Animal Research” section of this white paper).
Fund-subcontracted molecular biology support for immune function tests in human chiropractic studies. This funding would be available to all investigators examining chiropractic issues.
Specific immune function research topics.
The biomechanical spine lesion
Evaluate the potential adverse contributory effect of biomechanical spine lesions on specific immune function–related health conditions (using animal models with predispositions to conditions that mimic illnesses seen in humans such as: asthma, cancer, diabetes, inflammatory bowel disease, multiple sclerosis, and systemic lupus erythematosus).
Evaluate the putative adverse contributory effects of biomechanical spine lesions on the nonspecific (stress) response using animal models of spine fixation (eg, the External Link Model).
Evaluate the potential therapeutic effect of SM on specific immune function related health conditions (using animal models with fully expressed conditions that mimic illnesses seen in humans such as: asthma, cancer, diabetes, inflammatory bowel disease, multiple sclerosis, and systemic lupus erythematosus).
Evaluate the potential therapeutic effect of SM on the nonspecific (stress) response using animal models of spine fixation (eg, the External Link Model).
Evaluate the potential therapeutic effect of SM on the nonspecific (stress) response in human subjects.
Evaluate the potential therapeutic effect of SM on specific immune function related health conditions in humans having immune related health conditions such as: asthma, cancer, diabetes, inflammatory bowel disease, multiple sclerosis, and systemic lupus erythematosus.
An informed understanding of the structure and function of the autonomic nervous system is important to the effective practice of the individual chiropractic clinician and to the development of the discipline of chiropractic. At a superficial level, it is widely theorized that reflexes mediated via the autonomic nervous system are probably the primary mechanisms by which subluxation and adjustment may affect the function of the internal organs. That is not to say that they are the only mechanisms because other neurologic mechanisms and also nonneurologic mechanisms have the capacity to modify organ function in response to somatic stimulation. However, somatoautonomic reflexes provide what are thought to be the dominant forces.
It is therefore absolutely essential that the practitioner understand the potential relationships between somatic stimulation and autonomic function, that they are equipped to monitor autonomic function and diagnose autonomic dysfunction, and that they understand the implications that treatments at their disposal hold for somatoautonomic interactions. For the discipline of chiropractic, an understanding of autonomic structure and function is essential to the design, execution, and interpretation of clinical studies. Such understanding also provides the principal scientific rationale for the extension of chiropractic adjusting beyond musculoskeletal care.
Having defined what could be achieved with an intelligent understanding of the autonomic nervous system, and to consider what steps the chiropractic profession might wish to take with respect to this area of science, it would be worth examining the present status of our knowledge base. With particular reference to research of interest to the discipline of chiropractic, knowledge up to the mid 1990s was exquisitely synthesized in the landmark monograph of Sato et al  published in 1997. This volume particularly emphasized basic physiologic work, particularly animal experimentation, which formed the bulk of work in autonomic neuroscience up until that time. Consequently, only 3 of approximately 750 studies cited in the monograph by Sato et al made reference to the effects of spinal stimulation, and all of these were animal studies.
In the past quarter century, however, tremendous advances have been made in the noninvasive measurement of autonomic function, such that experiments which previously could only be performed in anesthetized animals can now be replicated and exceeded in human subjects. In addition, the appreciation of the role of the autonomic nervous system in such common and important diseases such as diabetes and chronic heart failure has led to enormous growth in clinically oriented research. This is manifested in the large numbers of research articles published and in such concrete indicators as the founding of the journal Clinical Autonomic Research and in the renewal of the Journal of the Autonomic Nervous System as Autonomic Neuroscience: Basic and Clinical. Since the publication of the last white paper, certainly, thousands of research articles of relevance to chiropractic have been published. The challenge for the chiropractic profession is in accessing, understanding, and applying this knowledge. A relatively small number of articles pertaining to autonomic neuroscience are directly concerned with chiropractic interventions. These article, which are most patently relevant to SM, will now be summarized.
Articles written by chiropractors and related to autonomic neuroscience have tended to focus on the use of SM in the relief of visceral complaints, on the reasonable assumption that any observed therapeutic effects are mediated by the autonomic nervous system. Logically, this research has focused on cardiovascular function, which is relatively easy to monitor, of profound clinical importance, and the primary target of short-term autonomic behavior. In addition, it is assumed to be relatively easy to extrapolate from measures of cardiovascular function to autonomic nervous system behavior.
Until a decade ago, virtually all discussion of autonomic involvement in chiropractic interventions was based on small studies — often case studies or series of patients. Indeed, with older articles in the health sciences in general, there may have been a tendency to mix large doses of hypothesis with rather miserly portions of original data. One article described a patient with premature ventricular contractions who was treated with SM.  Case reports [363-365] have also referred to positive effects in hypertension. Another article referred to 2 patients who received SM for the treatment of hypercholesterolemia. 
Several clinical studies have previously examined the effects of SM on blood pressure and heart rate in conscious humans to draw conclusions about physiologic regulation. Tran and Kirby [367, 368] reported that neither cervical SM (with the subject supine) nor upper thoracic SM (with the subject prone) had any effect on heart rate in small cohorts of healthy young adults. They did report small changes in systolic and diastolic blood pressure. However, these proposed effects are not supported by statistical analysis of their raw data — the authors only gave their impression of the general trend in blood pressure readings. An additional problem is that the authors did not give the time frame over which pre- and posttreatment measurements were taken, that is, seconds, minutes, or hours before and after treatment. Hood  reported that with chiropractic treatment and lifestyle modification, 75 patients who previously had high or low blood pressure, subsequently had readings closer to the norm. It is unclear from the written report whether the trial was prospective or retrospective, nor is it clear when blood pressure readings were taken in relation to treatment. Hence, the effects described may be nothing more than a demonstration of regression toward the mean.
McKnight and DeBoer  reported statistically significant decreases in systolic and diastolic blood pressure in a cohort of 53 healthy young college students (versus no significant changes in a control cohort of 22 subjects). The treatment applied was a cervical manipulation with the subject sitting. Blood pressure was measured by sphygmomanometry within the few minutes before treatment and within the 1 minute after treatment. On the other hand, Nansel et al  detected no responses in blood pressure and heart rate to sitting cervical SM in healthy young adults when comparing measures taken 15 minutes or more before treatment and 5 minutes or more after treatment. Yates et al  reported statistically significant decreases in systolic and diastolic blood pressure in a small cohort of mildly hypertensive patients receiving a single upper thoracic manipulation applied with a mechanical percussive device. Blood pressure was measured with a sphygmomanometer before and “immediately after” treatment. The effects of authentic manipulation could not be shown using placebo or sham treatments. However, there was no follow-up of patients beyond the day of treatment.
Knutson  reported on a cohort of patients (including normotensives and hypertensives) who received SM for “upper cervical joint dysfunction.” With blood pressure measured in the few minutes before and within 2 minutes after treatment, there was a significant decline in systolic, but not diastolic, blood pressure. There was no follow-up after the single treatment. In a small pilot study  examining the effects of SM vs massage in subjects with essential hypertension, small declines in systolic and diastolic pressure were seen over the course of several months, with the greatest effects achieved in the control (no treatment) group. No statistical comparison was made for effect size in this pilot study. A study by Goertz et al  showed that the addition of SM to dietary management did not improve outcomes (decreases in systolic and diastolic pressure at ~4 weeks) for subjects with “high-normal blood pressure or stage 1 hypertension.” Either the authors themselves or those who have subsequently interpreted these various studies have been inclined to speculate that increases in heart rate and blood pressure reflect increases in cardiac sympathetic output or, conversely, withdrawal of vagal output. In fact, better markers of autonomic function exist as do better markers of cardiovascular health.
None of these relatively few studies in healthy and hypertensive subjects report significant long-term effects on systolic or diastolic blood pressure. On the other hand, some studies have found statistically significant effects for cervical and thoracic manipulation in the few minutes immediately after treatment. This might have been predicted by reference to basic knowledge of autonomic function, which should have been referenced in designing these studies. An additional design problem with virtually all physiologic and clinical studies published until the late 1990s concerning SM and cardiovascular function lies in the methods of subject monitoring. In all instances cited, the investigators have relied on sphygmomanometry to measure blood pressure. In as much as this technique is inappropriate for continuous monitoring of blood pressure, and blood pressure fluctuates continuously in the conscious subject, most studies published so far on the short-term effects of SM on blood pressure have little credibility. Unlike conventional sphygmomanometry, arterial tonometry is appropriate for the continuous measurement of cardiovascular function during SM.  However, this rather expensive technology has found only limited application in chiropractic studies. 
Chiu and Wright  used skin conductance and skin temperature changes to infer responses of cutaneous sympathetic efferents to cervical mobilization. These results were consistent with cardiovascular effects reported separately.  This group has also looked at skin conductance and skin temperature responses to peripheral joint manipulation  and physiologic responses to cervical manipulation in patients with lateral epicondylalgia  and cervical pain.  This is an exceptional instance of a team of researchers progressively pursuing the neurophysiologic effects of manipulation, allowing them to refine their experimental technique and to develop robust hypotheses to explain their results.
Notwithstanding the successes of the University of Queensland team cited above, a technology that perhaps deserves more attention in investigations of SM involves the computation of cardiac autonomic output from electrocardiographic recordings. This technology is based on the understanding that heart rate is regulated on a beat-to-beat basis by a short feedback loop, which involves peripheral baroreceptors, brain stem nuclei and autonomic cardiac efferent fibers. An increase in the discharge rate of cardiovascular baroreceptors, as what occurs with increases in blood pressure, results in a reflex inhibition of sympathetic output to the cardiovascular system as well as an increase in cardiac vagal activity and, so, a decrease in heart rate. It is thought that the primary homeostatic purpose of this feedback system is to maintain mean arterial blood pressure around an optimal “set point,” allowing that the “set point” may vary in the long and short term, both in health and because of disease.  However, the increase in heart rate, which accompanies inspiration, probably also serves a role in increasing respiratory efficiency.  Noxious input and, so, activation of the periaqueductal gray matter tends to dampen baroreflexes.
Rhythmic oscillations in intrathoracic pressure, imposed by the respiratory cycle, modulate the normal fluctuations in blood pressure that result from the cardiac cycle. Hence, via the baroreflex, the respiratory rate entrains rhythmic variation of heart rate. This oscillation in heart rate, known as the respiratory sinus arrhythmia, is largely attributable to vagal (parasympathetic) activity in conscious humans at rest or during low levels of physical activity.  Somewhat slower intrinsic oscillations in sympathetic preganglionic neuron activity may result in subsequent superimposition of an additional low-frequency oscillation on blood pressure and heart rate. The result is a literally chaotic variation in cardiac rhythm resulting from the complex interaction of reflexes mediated by vagal and sympathetic cardiac efferents. The relative roles of the sympathetic and parasympathetic systems in heart rate variability (HRV) differ between species according to the access of different efferents to the myocardium and to the electrical conduction system of the heart. The physiologic contributions of the sympathetic and parasympathetic nerves and their anatomical details are of enormous importance in the surgical management of cardiac disease (see, eg, Ref. ), and probably have implications for responses to more benign somatic stimuli as well, including subluxation and adjustment.
Studies using HRV to derive measures of autonomic responses to SM and mobilization are beginning to be conducted. Only 1 article has reported a controlled study of the effects of SM on HRV.  This article showed that SM was associated with increases in cardiac sympathetic output and in the balance of sympathetic to parasympathetic flow to the heart, notwithstanding decreases in heart rate. Additional studies yet to be published corroborate these effects. These physiologic studies in well subjects have subsequently led to a pilot investigation in cardiac patients.
If one examines the modern history of investigations of the effects of SM on autonomic function, an ironic picture emerges. The earlier studies tended to focus on clinical outcomes from which conclusions concerning neurophysiology were extrapolated. Considering the results of more recent studies, these conclusions seem to have been largely mistaken. The technology used in earlier studies was largely inappropriate to the experimental questions being asked. In the past 10 years, there has been a shift to more appropriate technologies, which have been applied first in physiologic studies of well subjects and only later in studies of patients. These are, of course, very positive developments, although there is still a large gap between the cutting edge of technology (and knowledge) and investigations of SM.
Initiate a program to educate the educators, beginning with heads of programs, including academics in research, neurosciences, and chiropractic principles. The Research Agenda Conference meetings serve this objective in part. However, the concentrated nature of these meetings does not allow justice to be done to the wealth of possibilities that chiropractic could be exploring in neuroscience research.
A body, or bodies, should be entrusted to develop a current curriculum, which would serve as a common resource for all chiropractic educational programs. This is not to imply that all colleges should be teaching the same content. However, it is a significant challenge for some faculty to access reliable “chirocentric” resources in the neurosciences. There seems little reason to not have faculties cooperating on developing a shared resource.
A funding mechanism needs to be created to identify and support autonomic research of importance to chiropractic. Research funding decisions must be based first and foremost on scientific rigor.
The profession should create an endowed chair in autonomic research. This should be in an environment best equipped to produce research of the highest quality, to mentor young researchers, and to contribute to the common body of knowledge of the health sciences. This may possibly be in a chiropractic educational institution, but a university or free-standing research institution should also be considered.
The literature summaries of the 6 topic sections (anatomy, biomechanics, somatic nervous system, animal models, immune system, and human studies related to the autonomic nervous system) indicated that a significant body of basic science research evaluating chiropractic spinal adjusting has been completed and published since the 1997 basic science white paper. Much more basic science research in these fields needs to be accomplished, and the recommendations at the end of each topic section should help researchers, funding agencies, and other decision makers develop specific research priorities.