One possible reason for such differences might be the use of the low frequency (LF)/high frequency (HF) ratio as an indicator of ANS activity, where HF represents PNS efferent activity and LF corresponds to both PNS and SNS efferent activity. This method of assessing HRV has been criticised due to oversimplification of the complex non-linear interactions between the SNS and PNS (Billman 2013). More recently, Sampath et al. (2017), using a reliable measure (near-infrared spectroscopy) to assess SNS activity, reported an immediate sympathetic excitation following thoracic manipulation. Interestingly, this study also investigated pre and post manipulation HRV data but found no statistically significant difference between the groups. Nevertheless, the findings of this study need to be interpreted cautiously, as it was based on asymptomatic male subjects, and there has been a report of ANS dysregulation in chronic pain patients. Hence, more research on symptomatic population is warranted.
Effects of manipulation-induced autonomic changes on supraspinal mechanisms
As discussed above, there is a complex interaction between the ANS and the pain system, and the PSNS plays a significant role in modulating pain and inflammation. Hence, considering the evidence of immediate sympathetoexcitatory responses following manipulation, Sampath et al. (2015) suggested that these SNS changes might be linked to changes in pain-modulating supraspinal mechanisms. In support of this hypothesis, the authors cited two imaging studies (Ogura et al. 2011; Sparks et al. 2013) that demonstrated the effects of manipulation on several supraspinal structures including the cerebellar vermis, middle temporal gyrus, insular cortex, inferior prefrontal cortex and anterior cingulate cortex. All these structures have been reported to be involved in the regulation of autonomic function (Kenney and Ganta 2011). On the other hand, there has been a growing body of evidence in support of the manipulation-induced neural plastic changes (Daligadu et al. 2013; Lelic et al. 2016; Taylor and Murphy 2010) occurring in various brain structures such as the cerebellum, basal ganglia, prefrontal cortex, primary sensory cortex and primary motor cortex. Taken together, although there is no direct evidence in support of Sampath et al.’s (2015) hypothesis, this might be a fruitful area of research for future studies.
Co-activation of the neuroendocrine system
The hypothalamus region is known for coordinating stress responses by activating the hypothalamic–pituitary (HP) axis and a neural pathway involving the PSNS. The hypothalamic–pituitary–adrenal (HPA) axis has been considered to be the central stress response system and is known to release adrenal glucocorticoid (cortisol), which is a class of corticosteroids that are well recognised in the literature for their anti-inflammatory and immunosuppressive actions (Ulrich-Lai and Herman 2009). On the other hand, as discussed above, the SNS has been reported to serve as a mediator between the somatic and supportive processes. Hence, it has been well established that both the SNS and HPA axis could play a significant role in the modulation of acute and chronic inflammation, and the neuroendocrine (SNS–HPA axis) mechanisms are involved in the pain relief and tissue-healing processes (Chrousos 2009; Ulrich-Lai and Herman 2009). These two systems have also been reported to work together, overlapping the underlying neural circuitry (Chrousos 2009). In addition, the evidence suggests that spinal manipulation could influence the activity of both the SNS and HPA axis. Several studies have assessed the effect of spinal manipulation on the HPA axis, and an immediate increase in serum cortisol levels following manipulation has been observed in both symptomatic and asymptomatic patients (Padayachy et al. 2010; Plaza-Manzano et al. 2014).
Considering the above facts, Sampath et al. (2015) hypothesised that there could be an association between SNS changes and HPA axis responses, and post-manipulation changes in the SNS might be accompanied by HPA axis changes. The authors proposed possible neural reflex pathways in support of this hypothesis. They suggested that HVLAT at the thoracolumbar segment of the spine would result in excitation of the preganglionic sympathetic cells and subsequent stimulation of mechanoreceptors. These inputs would then travel to several regions of the brain stem and subsequently would lead to opioid-independent analgesia by influencing the hypothalamus and PAG (periaqueductal grey) matter in the midbrain. The hypothalamic release of the corticotropin-releasing factor (CRF) would then occur to modulate the SNS and HPA axis response. The neuroendocrine (SNS–HPA axis) system would then release its end products (catecholamines and glucocorticoids) to initiate anti-inflammatory and tissue-healing actions. However, to date, only one study (Sampath et al. 2017) has been conducted to investigate the SNS–HPA axis response to manipulation in the same trial. Although this study reported a reduction in salivary cortisol level immediately after thoracic manipulation and observed an immediate effect of manipulation on the SNS, the clinical relevance of such changes is so far unknown. Therefore, more research is needed to determine the true clinical significance of neuroendocrine response following manipulation.
Hypoalgesic effects
It is thought that four types of mechanism contribute to the hypoalgesic effects of spinal manipulation.
Segmental inhibition
The concept of this mechanism is based on Melzack and Wall’s (1965) gate control theory of pain. This theory proposes that nociceptive (small diameter) A-delta (A-δ) and C sensory fibres carry the pain stimuli to the dorsal horn and ‘open’ the substantia gelatinosa layer, whereas non-nociceptive (large diameter) A-β fibres inhibit the transmission of pain signals by blocking the entry of A-δ and C fibres. Because mechanical stimulus applied during spinal manipulation may alter peripheral sensory input from paraspinal tissues, it has been presumed that manipulation may influence the gate-closing mechanism by stimulating the A-β fibres from muscle spindles and facet joint mechanoreceptors (Potter et al. 2005). Systematic reviews by Millan et al. (2012) and Coronado et al. (2012) have critically reviewed studies that examined the hypoalgesic effects of spinal manipulation on experimentally induced pain. Most of the studies included in these two reviews observed a segmental hypoalgesic effect of manipulation, and suggested that supraspinal pathways might be involved in the segmental mechanism. In addition, the involvement of a segmental mechanism in the modulation of pain perception has also been proposed by numerous studies investigating the neuromuscular effects of spinal manipulation (see ‘Neuromuscular effects’). However, it needs to be determined whether the observed local hypoalgesic effect following manipulation is merely a reflex effect on the pre-existing painful condition itself or due to activation of the endogenous pain inhibitory system.
Activation of descending pain inhibitory pathways
This mechanism is based on the effects of spinal manipulation on the pain-modulatory neural circuitry. Manipulation has long been thought to modulate the non-opioid hypoalgesic system by activating the descending pain-modulation circuit, especially serotonin and noradrenaline pathways, from the PAG and rostral ventromedial medulla (RVM) of the brain stem (Pickar 2002; Vernon 2000; Wright 1995). This hypothesis has been supported by both animal model and human studies. In laboratory animal models (Reed et al. 2014; Skyba et al. 2003; Song et al. 2006), objective evidence has been found in support of a central antinociceptive effect that appeared to be mediated by serotoninergic and noradrenergic inhibitory pathways. The findings of human studies (Alonso-Perez et al. 2017; O’Neill, Ødegaard-Olsen and Søvde 2015; Sterling, Jull and Wright 2001) conducted on both symptomatic and asymptomatic subjects are also consistent with the findings of animal models. However, although human research supports a non-opioid form of manipulation-induced hypoalgesic effect through activation of some type of descending inhibitory mechanism, the exact circuit is yet not agreed upon. Because neural responses following spinal manipulation may vary depending on the rate of force application and the location at which the thrust is applied (Cambridge et al. 2012; Downie et al. 2010; Nougarou et al. 2016), it has been assumed that variations in mechanical parameters of manipulation may activate different descending inhibitory pathways (Savva, Giakas and Efstathiou 2014). Therefore, future research should be performed to investigate the exact descending pain-modulatory circuit involved after spinal manipulation, and these studies should also carefully consider the force/time and contact site characteristics of the given intervention.