Advanced Osteopathic and Chiropractic Techniques for Manual Therapists. Giles Gyer

the EMG signal are the two aspects that quantify muscle activity changes (Currie et al. 2016). Experimental studies done to assess neuromuscular responses to spinal manipulation reported both increases and decreases in EMG amplitude following manipulation (Bicalho et al. 2010; Ferreira, Ferreira and Hodges 2007; Lehman 2012; Lehman and McGill 2001). It is to be noted that most authors, including Lehman and McGill (2001), reported a reduction of paraspinal muscle activity following manipulation in resting phase. The conflicting results, however, appeared when EMG amplitudes were analysed during dynamic activity (flexion or extension). Nevertheless, most of the high-quality experiments published to date reported reduced paraspinal voluntary EMG amplitude during extension and relaxation phases (Lehman 2012). The changes in EMG amplitude in response to manipulation indicate that the underlying mechanism of spinal manipulation may involve the disruption of the pain–spasm–pain model.

      The timing of the EMG signal is another measure of the muscle activity changes. Muscle activity onset delay quantifies the reflex response of a given spinal manipulation. Onset delay of a muscle following manipulation is too short, and varies in a wide range, from 1 to 400 milliseconds (Colloca, Keller and Gunzburg 2003; Currie et al. 2016; Keller, Colloca and Gunzburg 2003); thus, it is unlikely to be activated voluntarily (Herzog 2000). On the other hand, because a spinal reflex is assumed to take place within 120 milliseconds (Wilder et al. 1996), there is a high likelihood that a spinal reflex response may be involved with the muscle activity onset delay. Furthermore, in a recent study Currie et al. (2016) quantified differences in muscle activity onset delay between symptomatic and asymptomatic participants following lumbar manipulation, and found that those with LBP (symptomatic) had longer onset delays than their healthy (asymptomatic) counterparts, although the difference in timing was only 5 milliseconds. The authors suggested that the delayed neuromuscular response in the symptomatic group in response to spinal manipulation might be due to the involvement of capsule mechanoreceptor pathways. In support of this claim, they cited Herzog’s (2000) work, where the author anticipated the faster activation of the muscle spindle pathways than capsular reflex pathways because of their reliance on large diameter Ia afferents.

      From the above discussion, it is evident that spinal manipulation results in neuromuscular responses, involves spinal reflex pathways and may reduce muscle hyperactivity. However, it needs to be investigated whether the evoked short-latency changes in EMG amplitude and timing following manipulation indicate a clinically significant outcome or merely a short-term effect.

      Modulation of gamma motor neuron activity

      Korr’s (1975) theory of the facilitated segment is a decades-old theory that has been used to interpret the mechanism of manipulation. From the early evidential basis, Korr hypothesised that a painful segment has a facilitatory response, and proposed that an increase in gamma motor neuron activity could lead to muscle hypertonicity by reflexly facilitating the alpha motor neuronal hyperexcitability. Korr suggested that spinal manipulation could calm the excited gamma motor neurons by increasing joint mobility, producing a barrage of proprioceptive afferent impulses. However, one major limitation of Korr’s theory is that it lacked the neural pathways (i.e., afferent input likely to arise and reflex pathways that may be activated due to spinal manipulation) for its proposed mechanism of action. Interestingly, the pain–spasm–pain cycle (Travell et al. 1942) sheds some light on the possible neural pathway that may be involved in gamma motor neuron excitability. Johansson and Sojka (1991) proposed that this neural pathway would involve a hyperactive spinal stretch reflex, which is a process that involves skeletal muscle contraction and is thought to occur when the muscle spindles and Ia afferents are activated due to stretching of the muscle (Trompetto et al. 2014). Johansson and Sojka (1991) postulated that nociceptive afferents directly project on the gamma motor neurons, which react by increasing the output of muscle spindles, allowing the associated afferent nerves to signal changes in muscle length. This, in turn, results in the hyperexcitability of alpha motor neurons and subsequently leads to increased muscle activation.

      As stated before, the pain–spasm–pain model is not unequivocally supported in the literature. Several authors have suggested that the sensitivity of muscle spindles is not affected by LBP, or paraspinal tissues do not undergo noxious stimulation (Birznieks, Burton and Macefield 2008; Zedka et al. 1999). Many studies still support the concept that spinal manipulation disrupts the pain–spasm–pain cycle and that it works by decreasing the hyperactivity of underlying nociceptors, consequently leading to stretch reflex attenuation and subsequent reduction in muscle activation (Herzog 2000; Pickar and Bolton 2012; Potter et al. 2005). Recently, however, two novel studies have established that with spinal manipulation, corticospinal or stretch reflex excitability can be attenuated. In the first study done to quantify the effects of spinal manipulation on stretch reflex excitability, Clark et al. (2011) observed an attenuation of stretch reflex of the erector spinae muscles when spinal manipulation produced an audible cracking sound. The authors suggested that manipulation might mechanistically act to reduce the output of muscle spindles and other segmental sites in the Ia reflex pathway. The second study was conducted by Fryer and Pearce (2012) on asymptomatic participants. The authors demonstrated a significant reduction in corticospinal and spinal reflex excitability following HVLAT manipulation that produced an audible cavitation. They also suggested that considerable alterations in corticospinal excitability could lead to changes in motor recruitment strategies.

      These findings provide more insight into the possible segmental mechanisms of spinal manipulation. In addition, because an increased stretch reflex gain forms the basis of one of the neural pathways of the pain–spasm–pain cycle, it can be said that spinal manipulation may function via the pain model by attenuating stretch reflex hyperactivity, consequently reducing the hyperexcitability of gamma motor neurons.

      Modulation of alpha motor neuron activity

      The involvement of alpha motor neurons in the modulation of musculoskeletal pain has been proposed by two of the prominent theories of pain: (1) the pain–spasm–pain cycle (Cooperstein, Young and Haneline 2013) and (2) the pain-adaptation model (Lund et al. 1991). The pain–spasm–pain model proposes two distinct neural pathways that contribute to pain. However, both theories have one common basis: that hyperexcitability of the alpha motor neuron pool leads to increased muscle activity. One neural pathway is described above (see ‘Modulation of gamma motor neuron activity’). Another pathway involves the projections of nociceptors onto alpha motor neurons via excitatory interneurons. On the other hand, the pain-adaptation model postulates that pain increases muscle activity when the muscle acts as antagonist and decreases it when active as agonist. The neural pathway proposed for this model involves feedback of nociceptive afferents projecting onto alpha motor neurons via both excitatory and inhibitory interneurons. The central nervous system (CNS) is thought to control the function of these interneurons and provide motor command of whether to excite or inhibit the alpha motor neuron pool (van Dieën et al. 2003). In short, regardless of the exact neural pathways, it may be said that the alpha motor neuron excitability forms the basis in the mechanism of musculoskeletal pain, as the modulation of alpha motor neurons correlates with changes in muscle activation.

      Spinal manipulation has been thought to relax or normalise hypertonic muscle through modulating alpha motor neuron activity. However, the exact effect(s) of manipulation on motor neurons is still unknown. As described above (see ‘Muscle activation’), most of the higher-quality EMG studies conducted to date demonstrated a significant attenuation of muscle activity following manipulation during forward bend or lying prone position (Lehman 2012). In a recent study on LBP patients, after observing reductions in EMG muscle activity during the flexion-relaxation phase, Bicalho et al. (2010) suggested that such decreases in EMG amplitude might be due to two different scenarios: (1) the hyperexcitability of the alpha motor neuron pool was decreased following spinal manipulation, or (2) manipulation increased the inhibition of the alpha motor unit. Nevertheless, the clinical relevance of EMG amplitude changes on the motor neuron pool is unclear, as EMG muscle activity changes were found to be transient in nature, and several studies have reported conflicting results.

      Two experimental techniques that have been used to effectively measure motor neuron activity after mechanical stimulation include the H-reflex and transcranial magnetic stimulation (TMS). The H-reflex technique assesses