Fossil evidence of trepanning suggests that man has used high threshold physical techniques in the treatment of disease since Neolithic times (Martin 2000, Parry 1936). Bone etchings from China dating back to 1600 bc are said to provide some of the earliest evidence of acupuncture techniques. Older still are the sharpened stones called Bian shi, although it is questioned whether or not these were actually instruments of acupuncture (Bai & Baron 2001). Harder evidence – in a softer format – comes from the silk scrolls found in Han Tomb No. 3 (dated to 168 bc) at Mawangdui, Changsha, China in the early 1970s. These manuscripts describe an early meridian system with 11 rather than 12 paired meridians and the use of moxibustion, which is a treatment involving the application of heat by burning the herb Artemisia vulgaris. The Pericardium meridian is missing (Chen 1997) from these early manuscripts. There is also an emphasis on information derived from tactile examination of the living body (Hsu 2005) rather than from dissection post-mortem. However, there is no description of acupuncture needling in these manuscripts (Bai & Baron 2001). The discovery of Ötzi, the Tyrolean iceman frozen from 3200 BC, suggests the use of a therapeutic needling technique, with a needle made from bone, which may have developed in Europe (Dorfer et al. 1999). It seems clear that acupuncture-like therapies have developed independently in different civilizations around the world and this is probably due to late evolutionary features in the mammalian nervous system, combined with intelligence, and the consequent use of tools in humans.

Children learn at a very early age to rub energetically directly over the site of acute pain to reduce the noxious sensation. In the case of a more chronic discomfort from aching, ‘knotted’ muscle, we tend to massage the local tissues more deeply and vigorously even though doing so may temporarily exacerbate the discomfort. This is likely to be conditioned behavior resulting from the analgesic effect of somatic sensory stimulation. With the development of stone tools it is easy to hypothesize a progression of therapeutic techniques which resulted ultimately in piercing the skin and muscle at a site of chronic pain. It may be that successful treatment of myofascial pain by piercing the body at the site of tenderness not only encouraged the practice, but also lead to the recognition of areas of the body which were most likely to harbor these tender points. In some parts of the world, people developed superficial techniques of scratching or cauterizing the skin, whereas in the Far and Middle East the technique of acupuncture developed (Cummings 2004).

Traditional acupuncture

The development of acupuncture points probably resulted from clinical observation that certain places in the body were more likely to harbor tender points than others and that treating these points by pressure or piercing could relieve pain and various other non-painful symptoms. Early physicians would have also noted that careful examination of the body surface revealed tender points in healthy subjects. Consistent patterns of pain referral from myofascial trigger points (TrPs) and the relief resulting from needling these and other muscle points would have lead them to make links between some of the points. Radiation patterns of painful medical conditions such as sciatica, other radiculopathies and possibly the consistent rashes of herpes zoster would have added to the impression that the established points were connected. These hypotheses do not explain the location of all acupuncture points, nor the paths of all the meridians, but there is clearly considerable overlap between myofascial trigger points and acupuncture points (Melzack et al. 1977), and between the pain referral patterns of the former and meridians (Dorsher 2009); although these potential correlations have caused great debate, and the theoretical backgrounds of these concepts are clearly distinct.

Acupuncture was probably used pragmatically by the Chinese and others for centuries before it became systematized within a documented form of medicine some 2000 years ago (Veith 1972). The theories developed were influenced by rational observations imposed upon a limited clinical knowledge base and in the philosophical framework of Taoism. The tendency towards syncretism resulted in the adoption and inclusion of many different theories, and over the centuries this has resulted in the development of a complex system of medicine. Whilst it can be initially unpalatable to the skeptical Western scientist, closer inspection reveals that Traditional Chinese Medicine is built on a series of logical assumptions, and although some of these are clearly wrong, many may still represent valid clinical observations.

Western medical acupuncture

Western medical acupuncture is a term with a variety of potential meanings. The most literal interpretation invokes thoughts of geographical boundaries, but the term was probably introduced to distinguish a developing system of needle therapy with a basis in Western medical science from its traditional philosophical roots which happened to be in the East. Filshie & Cummings (1999) interpret ‘Western Medical Acupuncture’ as the scientific application of acupuncture as a therapy following orthodox clinical diagnosis. It is important to note that the scientific evaluation of acupuncture is not restricted to the West (Han & Terenius 1982), and therefore adherence to a geographical definition is inappropriate. Probably a more accurate description of ‘Western Medical Acupuncture’ (WMA) is a modern scientific approach to therapy involving dry needling of tissues, which has been developed from the introduction and evaluation of traditional Chinese acupuncture techniques in the West (Cummings 2004).

More recently the definition of WMA has been reconsidered and redefined (White 2009):

Western medical acupuncture is a therapeutic modality involving the insertion of fine needles; it is an adaptation of Chinese acupuncture using current knowledge of anatomy, physiology and pathology, and the principles of evidence based medicine.

Modern scientific method was established by Galileo in the 17th century when he introduced systematic verification through planned experiments to the existing ancient methods of reasoning and deduction (MacLachlan 1999). This system was adopted by the scientific community throughout the globe, and with only the addition of statistical analysis it remains established practice today. The ethical practice of medicine requires the practitioner to understand and use scientific method. However, there is great debate over the use of certain methods of testing efficacy when applied to potentially complex interventions such as acupuncture.

Neurophysiological mechanisms of the technique

Neurophysiology of acupuncture needling

The therapeutic effects of acupuncture needling are mediated through stimulation of the peripheral nervous system, and so can be abolished by local anesthetic (Chiang et al. 1973, Dundee & Ghaly 1991). In particular, stimulation of Aδ or type III afferent nerve fibers has been implicated as the key component in producing analgesia (Chung et al. 1984). The therapeutic effects of needling can be divided into four categories based on the area influenced: local, segmental, heterosegmental and general.

Local effects

Local effects are mediated through antidromic stimulation of high threshold afferent nerves, in the same way as the ‘triple response’, first described by Professor Sir Thomas Lewis (Lewis 1927, Rous & Gilding 1930). Release of trophic and vasoactive neuropeptides including neuropeptide Y (NPY), calcitonin-gene-related-peptide (CGRP) and vasoactive-intestinal-peptide (VIP) has been demonstrated following acupuncture in patients with xerostomia (Dawidson et al. 1998a, 1998b). It is likely that the release of CGRP and VIP from peripheral nerves stimulated by needling results in enhanced circulation and wound healing in rats (Jansen et al. 1989a, 1989b), and equivalent sensory stimulation has proved effective in human patients (Lundeberg et al. 1988).

Increased circulation resulting from nerve stimulation is probably one of the most important local effects of acupuncture, and, in rats, it appears to be principally mediated by the release of CGRP (Sato et al. 2000). The effect of acupuncture on muscle blood flow, however, may not rely solely on nerve stimulation (Shinbara et al. 2008). Under normal circumstances in healthy human subjects, blood flow in muscle and skin is increased by needling local muscle points and less affected by needling skin (Sandberg et al. 2003). But this situation may be reversed if the subject is very sensitive, for example, in patients with fibromyalgia (Sandberg et al. 2004). The increase in muscle and skin blood flow following local needling of muscle in patients with work-related trapezius myalgia appears to be lower than in healthy subjects and this may reflect the degree of sympathetic activation and hypersensitivity of these patients (Sandberg et al. 2005).

Segmental effects

Through stimulation of high threshold ergoreceptors in muscle, needling can have a profound influence on sensory modulation within the dorsal horn at the relevant segmental level. C fiber pain transmission is inhibited via enkephalinergic interneurones in lamina II, the substantia gelatinosa. Bowsher (1998) reviews the basic science literature, which supports this mechanism, and White (1999) appraises experimental and clinical evidence. Segmental stimulation appears to have a more powerful effect than an equivalent stimulus from a distant segment in modulating pain (Chapman et al.1977, Lundeberg et al.1989, Zhao 2008), local autonomic activity (Sato et al. 1993) and itch (Lundeberg et al. 1987). Aδ or type III afferent nerve fibers can be stimulated by superficial needling as well as by needling deeper tissues, but it seems that segmental stimuli from the latter (usually muscle) have a more powerful effect (Lundeberg et al. 1987, 1989, Ceccherelli et al. 1998, Zhao 2008).

When treating somatic pain, including muscle pain, in the clinical setting it is difficult to differentiate between local and segmental effects of treatment, since local needling can mediate both effects. Segmental effects are easier to illustrate when local needling is not possible, e.g. in visceral complaints. Segmental electro-acupuncture under the name percutaneous tibial nerve stimulation has been shown to affect bladder function in patients with overactive bladder symptoms (Van Balken et al. 2001, 2003, Macdiarmid et al. 2010, Peters et al. 2010).

Visceral blood flow following acupuncture has also been studied, and whilst segmental effects appear to dominate (Stener-Victorin et al. 1996, 2003, 2004, 2006), non-segmental mechanisms are also apparent (Uchida & Hotta 2008).

Heterosegmental effects

Whilst segmental stimulation appears to be the more powerful effect, needling anywhere in the body can influence afferent processing throughout the spinal cord. The needle stimulus travels from the segment of origin to the ventral posterior lateral nucleus of the thalamus, and projects from there to the sensory cortex. Collaterals in the midbrain synapse in the periaqueductal grey (PAG), from where inhibitory fibers descend, via the nucleus raphe magnus, to influence afferent processing in the dorsal horn at every level of the spinal cord. Serotonin is the prominent neurotransmitter in the caudal stages of this descending pain pathway, and the fibers synapse with the enkephalinergic interneurones in lamina II. A second descending system from the PAG travels via the nucleus raphe gigantocellularis; its fibers are noradrenergic, and their influence is mediated directly on lamina II cells, rather than via enkephalinergic interneurones. Diffuse noxious inhibitory control (DNIC) is the term introduced by Le Bars et al. to define a third analgesic system, which is induced by a noxious stimulus anywhere in the body (Le Bars et al. 1979). Heterosegmental needling exerts influence through all three mechanisms to different degrees (Bowsher 1998, White 1999), and possibly through others, as yet undefined.

General effects

These are more difficult to define, and there is clearly some overlap with heterosegmental effects. The latter term is used here to denote effects mediated at every segment of the spinal cord, as opposed to effects mediated by humeral means or by influence on higher centers in the CNS controlling general responses. Acupuncture needling has proven efficacy in the treatment of nausea and vomiting (Lee & Done 2004, Lee & Fan 2009, Vickers 1996), and this effect is likely to be mediated centrally. There is a substantial body of work that indicates the importance of β-endorphin and other endogenous opioids in acupuncture analgesia (Han & Terenius 1982, Han 2004, 2010, Zhao 2008), and correlations have been identified between the endorphin releasing effect of acupuncture and that of prolonged exercise (Thoren et al. 1990). Further correlations in terms of neuropeptide release have been noted (Bucinskaite et al. 1996), and it has been suggested that chronic activation of opioid systems by exercise, or potentially by acupuncture, may mediate enhanced immunity, with decreased upper respiratory infections and protection against some forms of cancer (Jonsdottir 1999).

Functional magnetic resonance imaging (fMRI) studies indicate general effects on limbic structures (Hui et al. 2000), and indicate the importance of the nature of the needle stimulus in achieving this effect (Hui et al. 2007, 2009, 2010). Such effects may be important in pain as well as other conditions that affect general wellbeing.

Whilst target-directed expectation (Benedetti et al. 1999) may theoretically play a role in the mechanism of acupuncture under some circumstances, the effects of acupuncture do not appear to be explained entirely by expectation (Kong et al., 2009b, 2009a). In clinical practice, context driven effects are considered important (Finniss et al. 2010), but in this environment it is challenging to untangle the direct effects of acupuncture needling on central nervous system structures from the indirect effects related to the context of treatment.

Trigger point needling

The mechanism of action of direct needling in the deactivation of trigger points is undetermined. Despite the fact that a causal relationship has not been established between direct needling of trigger points and improvement in symptoms, a discussion of the potential mechanisms involved may still be useful in developing future research questions. Simons et al. (1999) commented on the results of two trials which compare direct dry and direct wet needling of trigger points (Skootsky et al. 1989, Hong 1994) and conclude that the critical therapeutic factor in both techniques is mechanical disruption by the needle. The common factor is certainly needle insertion into the trigger point; however, Hong (1994) highlighted the importance of stimulating a local twitch response in achieving an immediate effect and with Simons cites evidence that the local twitch response is mediated by a segmental spinal reflex (Hong & Simons 1998). Fine et al. (1988) performed a rigorous experimental study in which trigger points were subject to direct wet needling, and clearly demonstrated that an opioid mechanism was involved in trigger point pain relief. In light of this evidence it seems likely that the needle works more often through sensory stimulation than through mechanical disruption, and this would be consistent with the mechanism of action of acupuncture analgesia (Han & Terenius 1982, Han 2004, 2010, Zhao 2008). Having said that, techniques vary considerably, and it is possible that the more vigorous and fast insertion trigger point needling has a direct mechanical effect on endplates, muscle spindles, or fibers themselves (see Chapter 2 for physiological mechanisms of trigger point dry needling).

Clinical research

Methodological difficulties of clinical acupuncture research

The principal methodological difficulties in clinical trials that study the efficacy of acupuncture are concerned with controls and blinding (Lewith & Vincent 1998, White et al. 2001b). For a placebo control to be credible the subjects receiving it must believe that they have had an active treatment, identical to, or at least equivalent in potency to, the active intervention. Ideally, for any needling therapy, the control should involve an inactive form of needling, but it seems clear that a needle placed anywhere in the body is likely to have some neurophysiological effect (Lewith & Machin 1983), perhaps as a result of the noxious stimulus (Le Bars et al. 1979) of a needle piercing skin, or perhaps related to context-driven and interactional effects including target-directed expectation (Benedetti et al. 1999) and complex conditioned responses (Lundeberg & Lund 2008).

An innovation in needle design (Streitberger & Kleinhenz 1998, Kleinhenz et al. 1999) appeared at first to overcome the problem of needle penetration of skin by using a blunt needle that slid up into the coiled metal of the handle. This device was credible to the subject, but in order to simulate needle retention in the body it needed to be attached to the skin. This was done by inserting it through an adhesive plaster dressing over a small plastic ring placed over the point. In practice, however, the blunt needles pushed with enough force to get through the plaster also occasionally penetrated the skin surface (Konrad Streitberger: personal communication 2001). The Park Sham Device, which consists of a plastic guide tube of adjustable height with a sticky base, was developed as an alternative method of holding the sham needle in place (Park et al. 2002), but the subject could be unmasked if a needle fell out of the device. A convincing control procedure should result in blinding of the subject, but it is almost impossible to blind an experienced therapist who is performing both real and sham needling techniques. A common way of reducing bias in this situation is to use a blind assessor. A non-penetrating needle device that blinds the practitioner as well as the subject has been developed and validated (Takakura & Yajima 2007, 2008); however, it seems that simple non-penetrating sham acupuncture procedures, such as blunted cocktail sticks tapped on the skin, can be highly effective in clinical trials (Cherkin et al. 2009), and so the measured efficacy of true acupuncture over sham techniques in clinical trials is often small and not statistically significant.