While the nature of the “meridian” or the “channel” was still in question, one may ask whether the nervous system or chemical mediators were involved. The results obtained in the human study were so straightforward that the analgesic effect could be totally prevented when the local anesthetic procaine was infiltrated into the deeper structures under the point, but not by its subcutaneous injection. The results suggest that it is the nervous tissue in the muscle and tendon that senses the stimulation. It was later made clear that the small-myelinated nerve fibers (A_β fibers and a small part of the A_δ fibers) are responsible for the transmission of afferent impulses to the spinal cord [8].

Another important step made in the study of the mechanisms of acupuncture analgesia was the cerebrospinal fluid (CSF) cross-perfusion study [9]. In order to test the hypothesis of whether there are chemical mediators produced in the brain that may be responsible for the analgesic effect, stainless steel cannulae were implanted into the lateral ventricle of the rabbit so that the brain ventricle can be perfused with artificial CSF, and the perfusate was then infused immediately to the cerebroventricle of the recipient rabbit. When acupuncture was administered to the donor rabbit, the pain threshold increased dramatically. During this period, the CSF was drawn from the donor rabbit and injected into the brain of the recipient. A significant increase of the pain threshold was observed in the recipient rabbit (Fig. 10.2), although no acupuncture was given to this animal. These results suggested that during the acupuncture, some chemical substance(s) with analgesic potency might have been produced, which can be removed from the donor rabbit to the recipient. This finding triggered the interest to explore the neurochemical mechanisms of acupuncture effects.

A literature search revealed serotonin, or 5-hydroxytryptamine (5-HT), to be a candidate for the mediators of the analgesic effect. Studies performed in rats and rabbits showed that increase of the availability of 5-HT in brain or spinal cord potentiated acupuncture analgesia, whereas blockade of 5-HT synthesis or receptor activation resulted in a significant decrease of the analgesic effect. All of the results pointed to the conclusion that 5-HT in the central nervous system plays an important role in the mediation of acupuncture analgesia [10].

In contrast to the unique effect of 5-HT in the entire central nervous system, the role played by norepinephrine (NE) was much more complicated. Most of the information suggested that NE in the spinal cord played a facilitatory role for acupuncture analgesia, in contrast to the antagonistic role in the brain [11].

The discovery of enkephalins in the pig brain in 1975 triggered a huge storm in the biomedical field. Every researcher in this field tried to find some relation with endogenous opioid peptides, and there was no exception for researchers of acupuncture analgesia. David Mayor [12] was the first to step into this field. He used the opioid receptor antagonist naloxone as the research tool and found that the analgesic effect of acupuncture for dental pain can be prevented by the subcutaneous injection of naloxone, suggesting the involvement of endogenous opioid substances. Since opioid receptors can be divided into three types, μ, δ, and κ, and naloxone is a nonspecific antagonist for all three kinds of opioid receptors, this pharmacological tool can hardly be used to make a further receptor-type differentiation. Using a specific antagonist for the three types of opioid receptor, Han and colleagues were able to find that the analgesic effect of 2-Hz stimulation is mediated by μ and δ receptor, whereas at 100 Hz, the effect is mediated by κ receptors [13]. Further studies using radioimmunoassay revealed that 2 Hz increased the release of enkephalins and endorphins in the CNS to interact with μ and δ receptors, whereas 100 Hz increased the release of dynorphin in the spinal cord to interact with κ receptors (Fig. 10.3) [13].

An interesting question was that if low- or high-frequency stimulation can only accelerate the release of a fraction of the opioid peptide family, can we design a pattern of frequency which can accelerate the release of all four kinds of opioid peptides. This may have a practical impact since there are reports showing that simultaneous activation of two types of opioid receptors may cause a synergistic effect [14, 15]. After a series of exhausting experiments performed in the rat, it was revealed that a frequency automatically alternating between 2 and 100 Hz, each lasting for 3 s, produced a significantly more potent analgesic effect than pure low or pure high frequency alone [13]. This is reasonable since a fraction of the enkephalins released during the low-frequency period may survive to the next period of high-frequency stimulation when dynorphin is released. The coexisting enkephalin and dynorphin may interact at the receptor sites to produce a synergistic effect.

In performing animal experiments of acupuncture analgesia, two basic phenomena called our attention. One is the marked individual variation or the unpredictability of the acupuncture effect, and the other is the gradual fading of the analgesic effect with time if acupuncture is administered too often in a short period of time. As a general rule, acupuncture analgesia needs time (about 30 min) to build up to its full potential, and the effect would decay when the acupuncture needle is poured off (Fig. 10.1) or left unattended (Fig. 10.4). If EA is given 30 min/h, for 4–5 h, the analgesic effect would decrease gradually (Fig. 10.5) [16]. This is not due to the local tissue damage caused by repeated needle insertion and manipulation, since the situation would remain even if the needle is inserted into a new point of the body without any tissue damage.

In searching for the possible mechanisms, a hypothesis was raised that, according to the concept of Yin and Yang balance in the Chinese philosophy, the existence of a natural pain-killing substance (endorphin) might be accompanied by the existence of another substance with an antagonistic effect (anti-opioid substance). After a careful survey, this putative anti-opioid substance was identified as cholecystokinin octapeptide (CCK-8) [17]. In fact, repeated EA produced an increase of the production and release of opioid peptides, and in the same time, there is a gradual increase of the CCK-8 in the central nervous system which plays an antagonistic role against opioids, thereby reduces the effect of EA. This phenomenon is termed “acupuncture tolerance,” to mimic the situation of “morphine tolerance” produced by repeated injection of morphine. It was interesting to find that for rats with CCK predominates over opioid peptides in the central nervous system, they can be a nonresponder toward acupuncture, or a weak responder but quickly developing into acupuncture tolerance. The situation can be reversed if (a) CCK antagonist is injected intracerebroventricularly or intrathecally to the rat, or (b) the gene expression of CCK is blocked by the antisense probe against preproCCK administered centrally. In that case, the nonresponder of acupuncture analgesia can be changed into responder and the diminished analgesic effect can be revived [17, 18]. The lesson learned from this mechanism is that acupuncture should not be given too often, or last too long in one session.

From neurophysiological point of view, acupuncture analgesia can be taken as a reflex action. The afferent comes from the nerve fibers (mostly Aβ fibers) innervating the acupoint, and the efferent is the descending pathway modulating the sensitivity of the dorsal horn neurons not only in the same segment but also in heterogenous segments. Studies in the rat revealed that 100-Hz stimulation of the acupoint would trigger the release of dynorphin in the spinal cord. After the destruction of the parabrachial nucleus of the brain stem, high-frequency EA would no longer produce an analgesic effect [19]. Conversely, 2-Hz EA induces the release of β-endorphin in the brain and enkephalin in the whole central nervous system. After the destruction of the arcuate nucleus of the hypothalamus (where β-endorphin neurons aggregated), 2-Hz EA would no longer elicit analgesic effect. Taken together, a diagram could be constructed to show the hypothetical neural pathway for acupuncture analgesia (Fig. 10.6). Neither low- nor high-frequency EA would work if a lesion is placed at the periaqueductal gray (PAG) of the midbrain [19].

Other neural pathways have also been proposed. For example, 100-Hz stimulation can evoke supraspinal long-term depression not only in normal rats [20] but also in sham operated rats subject to neuropathic pain [21], contributing to the mechanisms of high-frequency EA-induced analgesic effect.

A hypothetical diagram was proposed by Han, which gives a general picture of the neural network underlying acupuncture analgesia, at least for the control of the acute pain [6].

Functional magnetic resonance imaging (fMRI) was used to characterize the possible brain areas being involved in mediating the acupuncture effect. Since stimulation of any of the body sites would cause extensive changes in the brain MRI picture, it is hard to characterize the brain sites responsible for acupuncture-induced analgesic effect. Zhang et al. [22] tried to correlate the magnitude of acupuncture-induced BOLD signal change observed in identified brain area with the magnitude of the analgesic effect induced by 2- or 100-Hz EA stimulation,respectively. The results showed that the analgesic effect induced by low and high frequencies seems to be mediated by different, though partially overlapping brain networks. In either frequency, the averaged fMRI activation levels of bilateral secondary somatosensory area and insula, contralateral anterior cingulate cortex, and thalamus were positively correlated with the EA-induced analgesic effect. In the 2-Hz EA group, positive correlation was observed only in contralateral primary and supplementary motor areas, while negative correlation was observed in bilateral hippocampi. In 100-Hz EA group, positive correlations were observed in contralateral inferior parietal lobule and ipsilateral anterior cingulate cortex, while negative correlation was found in contralateral amygdale. These results suggest that functional activation of certain brain areas might be correlated with the effect of EA analgesia in a frequency-dependent manner. More work is needed in order to figure out the complicated neural network controlling acupuncture-induced analgesic effect.

In clinical practice, various kinds of methods have been used to secure the optimal stimulation of the acupoint. Manual needling (MA) is the classical technique, with the characteristics of preciseness, capable of directing the needle in various angles to find the maximal deqi sensation and freely adjusting the needle movement to obtain a specific effect termed “warm,” “cold,” etc.

Studies show that various kinds of manual needling produce different pattern of afferent impulses in the sensory nerves and activation of the dorsal horn neurons. Out of 12 different kinds of needle manipulation, Wang et al. [23] studied three most popularly used modes, that is, twist, drag-plug, and gradual mode, respectively. Single-unit recordings of the dorsal horn neuron of the rat showed clearly different patterns in the inter-spike intervals (ISI). Based on the reconstructed phase space, they analyzed the spatiotemporal behavior of the time series. The largest Lyapunov exponent, which is an important parameter for describing the nonlinear system behavior, varies significantly in different modes of acupuncture. However, various types of manual needling technique (e.g., “burning mountain” for hot, “frozen sky” for cold) need years to learn and to master.

Compared to manual needling, electroacupuncture (EA) is a modern approach based on the basic finding that the effect of acupuncture is relied on the integrity of the nervous system [7] and that delivering specific forms of electric impulses is the easiest way to activate the afferent nerves in a predictable manner, with the added advantage of time saving and very high reproducibility. Aside from the great time saving and the reproducibility of the treatment, a significant advantage of using EA is that you can try, in certain degree, to change the internal environment of the central nervous system according to the ever changing need of the body system, for example, the use of low frequency (2–4 Hz) for the production and release of enkephalins in the central nervous system and high frequency (80–120 Hz) for dynorphins in the spinal cord [13]. The clinical effects produced by EA of different frequencies can be very different. Study shows that for treatment of rat model of neuropathic pain produced by lumbar nerve ligation, 2 Hz is much effective than 100 Hz, with the involvement of mu opioid receptors [24]. In contrast, for the treatment of patients with spinal cord injury-induced muscle spasm, it is only 100 Hz, but not 2 Hz, which works [25], with the involvement of kappa opioid receptors [26]. In these extreme cases, one frequency may serve as the control of the other. Here, the credibility of the design for “control” is nearly perfect, since no one knows which frequency is better, even for the care provider. This frequency-specific design can be served as an example to show the specificity of the EA treatment, rather than a design of using a nonspecific skin touch for psychological “believing.” However, the frequency specificity does not apply for every disease. For example, for the treatment of rat model of complete adjuvant-induced arthritis, both high- and low-frequency EA work at a similar efficacy [27].

Unlike electroacupuncture (EA) which uses percutaneous (invasive) approach, the transcutaneous electric acupoint stimulation (TEAS) is a noninvasive way of stimulating the acupoint by the use of skin pads placed on the skin surface overlying the acupoint in lieu of the needle. This is also called “acupuncture without a needle.” Since the skin electrode is usually 4  ×  4 cm in size, it would never miss the “acupoint.” Since all the parameters are shown on the LED screen precisely, it can also be used by the patient or family under the instruction of the acupuncturist or the physician, thereby reduces the number of visit to the doctor.

The efficacy of EA in pain control has repeatedly been shown to be no less than manual needling. Wang et al. [28] had done a careful study in the rat experiment, comparing the analgesic effect produced by EA or TEAS, with the conclusion that TEAS is at least as effective, if not more effective, than EA. The analgesic effect produced by either method can be blocked by naloxone at the same degree, suggesting a similar underlying mechanism of action. Given that these forms of acupoint stimulation may have similar therapeutic effect and underlying mechanisms, we will make clear statement separately for MA, EA, and TEAS in the following text when clinical applications are to be mentioned.

In the recent literature, there is another term called “percutaneous electrical nerve stimulation (PENS)” in contrast to “transcutaneous electrical nerve stimulation (TENS)” [29]. In a commentary put forwarded by Cummings [30], the author mentioned that “PENS is neither different in principle nor in practice from EA. While the term accurately reflects the nature of the treatment, there is no substantial justification for referring to PENS as a novel therapy.”