Last July the main findings from my PhD study into the mechanism of spinal manipulation were published and, especially as an early career researcher, I’ve been delighted at the interest shown in the paper. Really this is testament to the benefits of open access publishing (although there are criticisms of this publishing model) where research papers are made freely accessible to anyone (with an internet connection) anywhere in the world as well as the channels for spreading the word provided by social media. Of course, as well as that, the topic of the mechanism of spinal manipulation continues to generate interest in the manual therapy professions as we continue to try (and try) to understand how it works.
I’d like to take this opportunity to shed some light on what the findings from the study might mean for practising clinicians by addressing questions/comments that have been directed to me that have appeared in social media. By necessity some of the questions/comments are paraphrased and I have tried hard not to misinterpret these. I’m only too happy to (try to) answer any further questions you might have as a result of reading these posts. Please post comments, especially if you disagree! (Although please be polite! It’s fine to attack methods and methodology, not people). The limitations of the study should always be kept in mind when interpreting the ‘clinical implications’ responses to each comment. I look forward to hearing from you! (and look out for Part 2 of this blog-post, coming soon).
Comment 1 – ‘The findings of the study make sense since SMT does not lead to plastic deformation of joint and disc and therefore cannot increase movement more than anatomical range of motion (not lead to hypermobility)’.
How should hypermobility be defined within the context of inter-vertebral motion? (As distinct from say, generalised hypermobility diagnosed by the Beighton scale?). It’s standard in the biomechanics literature to define it as motion at or above the 98th percentile of the curve describing the distribution of inter-vertebral motion in the population (study sample) as illustrated in the graph below:
Figure 1: Theoretical framework for the categorisation of inter-vertebral motion (Deitz et al. 2011)
In the study there were 20 segments that increased in range above the MDC (minimum detectable change) after four weeks of SMT; motion changes after SMT are more likely to be ‘true’ changes if larger than the MDC. (MDC = repeatability coefficient. This was calculated from the age/gender-matched healthy volunteer group who did not receive SMT. The MDC provided an estimate of how much we might expect inter-vertebral motion to vary between baseline and follow-up 95% of the time; in other words, how much inter-vertebral motion might vary in people with no neck pain not receiving SMT). The graph below shows the angular range classification (i.e. hypomobile, normal or hypermobile) of these 20 segments at baseline (blue) and at four-week follow-up (red) after eight treatment visits for SMT.
Figure 2: Baseline and follow-up angular range classification of segments that increased in range (n=20 segments)
You might note the red column at the far right-hand side of the graph which represents seven segments (in four patients) that were classified as hypermobile (>mean +2 standard deviations) at follow-up. These segments were classed as within normal range at baseline. Since this study was observational (no randomisation or true control group) it cannot be said that the SMT caused these segments to increase beyond the normal range, but these findings open up the possibility.
Clinical Implications? – The continued use of cervical manipulation in the absence of pain, or for very low levels of pain, might not be clinically justified since there appears to be the possibility of inducing segmental hypermobility…
Comment 2 – ‘SMT works by neurophysiological mechanisms’
One might reasonably consider this, especially in light of the findings from this study (in brief, patients tended to get better whether their inter-vertebral motion increased, decreased or did not change, and only one segment that increased in range was hypomobile at baseline). However, which neurophysiological mechanism(s)? As the model proposed by Bialosky et al (2009) below suggests, answering this is not going to be straightforward:
Figure key: The model suggests a transient, mechanical stimulus to the tissue produces a chain of neurophysiological [and psychological] effects. Solid arrows denote a direct mediating effect. Broken arrows denote an associative relationship which may include an association between a construct and its measure. Bold boxes indicate the measurement of a construct.
Figure 3: Model of proposed mechanisms of manual therapy (Bialosky et al. 2009) (If model difficult to see please click here to see the original)
Simply put, this model proposes that a mechanical stimulus, like SMT, sets off a chain of neurophysiological responses that are ultimately responsible for the clinical outcomes. A number of the mechanisms postulated by this model have been investigated, but mostly theorised, in relation to SMT. Most investigative research has used cadavers, animal models or asymptomatic volunteers which can provide useful information to inform patient studies but the findings from such studies are in themselves not immediately clinical useful. Further, when mechanisms are explored in patients they are rarely considered in relation to symptomatic changes (one of the strengths of my PhD study) and study designs rarely include a control group (one of the weaknesses of my PhD study). I’ll briefly take each neurophysiological mechanism that has been investigated in turn:
Descending pain inhibition?
One paper that was referred to in social media when discussing the findings from my study was the systematic review and meta-analysis by Coronado et al. (2012) presumably since the authors of the review concluded that SMT had a greater effect on increasing pressure-pain thresholds (PPT) compared to other interventions, suggesting an influence on central descending pain inhibition (Coronado et al. 2012). However, the conclusions of this review need to be interpreted with the following in mind: most of the studies included had been carried out on asymptomatic participants, there was a lack of studies linking changes in pain sensitivity to changes in clinical outcomes, and most studies only assessed short-term or immediate PPT changes (Coronado et al. 2012). Finally, this systematic review appears not to have taken account of the MDC (minimum detectable change) in assessing changes in PPT – if changes are not greater than MDC, it is difficult to be sure they are true changes.
In a study that assessed immediate changes in patients with neck pain after cervical or thoracic manipulation, while pain and PPT improved significantly, changes in PPT did not exceed the MDC (Martinez-Segura et al. 2012).
A review paper on SMT somatosensory activation reveals that this is an area of investigation that is still at the experimental stage (Pickar and Bolton 2012) so no firm conclusions can yet be drawn.
Effects on sympathetic nervous system?
A recent systematic review of research into the effects of spinal mobilisation on the sympathetic nervous system found seven randomised controlled trials that the authors rated as high quality. These studies found consistent increases in sympathetic nervous system activity across all outcome measures, indicative of sympathetic excitation, irrespective of the segments mobilised (Kingston et al. 2014). However, only one study evaluated changes in a symptomatic population and, since changes were not linked to outcomes, the clinical utility of changes in skin conductance, decrease in skin temperature, and especially of increases in respiratory rate, blood pressure and heart rate, are unknown and questionable (Kingston et al. 2014).
Effects on endocrine system?
Despite the role it plays in pain modulation, there has been little research of the effects of SMT on the endocrine system. In a small prospective case series (n=9, assumed to be asymptomatic) serum cortisol levels were not significantly different after four treatment visits for SMT (region of spine not stated) (Tuchin 1998). A second study compared salivary cortisol levels in a cervical SMT group, a sham group and a control group before and after treatment, and found no differences in cortisol changes between the (asymptomatic) groups (Whelan et al. 2002). Finally, despite the author’s tenuous claims to the contrary, Padayachy et al (2010) similarly found no differences in serum cortisol levels five minutes after lumbar manipulation in 30 asymptomatics (Padayachy et al. 2010).
Effect on inflammation?
In a cross-sectional study patients (n=27) with chronic and recurrent neck pain were found to have significantly higher levels of serum inflammatory mediators compared to controls with no neck pain (Teodorczyk-Injeyan et al. 2011). The presence of an inflamed joint is generally considered a contra-indication to manipulation, at least in the case of spondyloarthopathies (Assendelft et al. 1996). However, in an earlier study the same research group found that a single thoracic manipulation was associated with a greater decrease in inflammatory cytokines compared to sham or venepuncture controls (Teodorczyk-Injeyan et al. 2006). While experiments with an animal model suggest SMT might decrease inflammation associated with the inter-vertebral foramen (Song et al. 2006), the mechanism behind these decreases in inflammation, the duration of action and the importance of this in mediating patients’ clinical outcomes in neck pain or any other condition remains unknown.
Effects on the brain?
The use of functional MRI to monitor brain changes in response to lumbar mobilisation has been investigated recently (Meier et al. 2014). This appears to be feasible and might be a more promising avenue for investigating the neurophysiological effects of SMT. A promising future?
Clinical Implications? – SMT might work via neurophysiological mechanisms, but we don’t yet know which one(s) are most important i.e. those that lead to clinical outcomes. More research is required, using more robust methodology and in patient populations.
“Findings from study on the mechanism of spinal manipulation: Clinical Implications? Part 2” coming soon!