Acupuncture, the limbic system, and the anticorrelated networks of the brain
The study of the mechanism of acupuncture action was revolutionized by the use of functional magnetic resonance imaging (fMRI). Over the past decade, our fMRI studies of healthy subjects have contributed substantially to elucidating the central effect of acupuncture on the human brain. These studies have shown that acupuncture stimulation, when associated with sensations comprising deqi, evokes deactivation of a limbic-paralimbic-neocortical network, which encompasses the limbic system, as well as activation of somatosensory brain regions. These networks closely match the default mode network and the anti-correlated task-positive network described in the literature. We have also shown that the effect of acupuncture on the brain is integrated at multiple levels, down to the brainstem and cerebellum. Our studies support the hypothesis that the effect of acupuncture on the brain goes beyond the effect of attention on the default mode network or the somatosensory stimulation of acupuncture needling. The amygdala and hypothalamus, in particular, show decreased activation during acupuncture stimulation that is not commonly associated with default mode network activity. At the same time, our research shows that acupuncture stimulation needs to be done carefully, limiting stimulation when the resulting sensations are very strong or when sharp pain is elicited. When acupuncture induced sharp pain, our studies show that the deactivation was attenuated or reversed in direction. Our results suggest that acupuncture mobilizes the functionally anti-correlated networks of the brain to mediate its actions, and that the effect is dependent on the psychophysical response. In this work we also discuss multiple avenues of future research, including the role of neurotransmitters, the effect of different acupuncture techniques, and the potential clinical application of our research findings to disease states including chronic pain, major depression, schizophrenia, autism, and Alzheimer's disease.
Acupuncture, the limbic system, and the anticorrelated networks of the brain
Kathleen K.S. Hui
, Ovidiu Marina
, Jing Liu
, Bruce R. Rosen
, Kenneth K. Kwong
Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital & Harvard Medical School, Charlestown, Massachusetts, United States
Transitional Year Program, William Beaumont Hospital, Royal Oak, Michigan, United States
Default mode network
The study of the mechanism of acupuncture action was revolutionized by the use of functional magnetic
resonance imaging (fMRI). Over the past decade, our fMRI studies of healthy subjects have contributed
substantially to elucidating the central effect of acupuncture on the human brain. These studies have shown
that acupuncture stimulation, when associated with sensations comprising deqi, evokes deactivation of a
limbic–paralimbic–neocortical network, which encompasses the limbic system, as well as activation of
somatosensory brain regions. These networks closely match the default mode network and the anti-
correlated task-positive network described in the literature. We have also shown that the effect of
acupuncture on the brain is integrated at multiple levels, down to the brainstem and cerebellum. Our studies
support the hypothesis that the effect of acupuncture on the brain goes beyond the effect of attention on the
default mode network or the somatosensory stimulation of acupuncture needling. The amygdala and
hypothalamus, in particular, show decreased activation during acupuncture stimulation that is not
commonly associated with default mode network activity. At the same time, our research shows that
acupuncture stimulation needs to be done carefully, limiting stimulation when the resulting sensations are
very strong or when sharp pain is elicited. When acupuncture induced sharp pain, our studies show that the
deactivation was attenuated or reversed in direction. Our results suggest that acupuncture mobilizes the
functionally anti-correlated networks of the brain to mediate its actions, and that the effect is dependent on
the psychophysical response. In this work we also discuss multiple avenues of future research, including the
role of neurotransmitters, the effect of different acupuncture techniques, and the potential clinical
application of our research ﬁndings to disease states including chronic pain, major depression, schizophrenia,
autism, and Alzheimer's disease.
Published by Elsevier B.V.
The limbic system is a group of limbic, paralimbic and neocortical
brain regions that together play a concerted role in the regulation and
integration of cognition, affect, sensory perception, biological behav-
ior, and autonomic, immunological and endocrine functions. The
advent of functional magnetic resonance imaging (fMRI) enabled in
vivo investigation of brain function and deﬁnition of additional brain
networks. The activity of the resting brain, in particular, when
challenged with a task, appears to be organized into two anti-
correlated networks that regulate each other to maintain balance
(Fransson, 2005; Fox et al., 2005). These networks are termed the
task-positive network, which shows activation during a task relative
to rest, and the task-negative network, which shows deactivation
during a task relative to rest. The regions comprising these networks
The default mode network is an instance of a task-negative
network showing extensive deactivation when an attention-demand-
ing task is engaged. It is described in the literature as comprising of
clusters of regions in the medial prefrontal cortex, posterior medial
parietal cortex and medial temporal lobe that are highly active in the
awake and conscious resting state but become deactivated when
exposed to external stimuli such as cognition and conceptual tasks
(Binder et al., 1999; Buckner et al., 2008; Fransson, 2005; Golland
et al., 2008; Gusnard and Raichle, 2001; Shulman et al., 1997). The
task-positive network is comprised of the sensorimotor and atten-
tion-related cortices that become activated during goal-directed tasks
(Corbetta and Shulman, 2002). Although not as extensively engaged
as in the task-negative system, a few paralimbic structures such as the
anterior middle cingulate, right insula and dorsal division of the
posterior cingulate Brodmann area 23 constitute core regions in the
Our fMRI studies of the effect of acupuncture on the brain in
normal human subjects have led us to deﬁne a task-negative network
Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
⁎ Corresponding author. Athinoula A. Martinos Center for Biomedical Imaging,
Department of Radiology, Massachusetts General Hospital, 149 13th St., Charlestown,
MA 02129, United States. Tel.: +1 617 724 7194; fax: +1 617 726 7422.
E-mail address: firstname.lastname@example.org (K.K.S. Hui).
1566-0702/$ – see front matter. Published by Elsevier B.V.
Contents lists available at ScienceDirect
Autonomic Neuroscience: Basic and Clinical
journal homepage: www.elsevier.com/locate/autneu
for acupuncture that is centered on the limbic system. We have
named this network the limbic–paralimbic–neocortical network (Hui
et al., 2005; Fang et al., 2008; Wang et al., 2007). Inspection of the
patterns of response of this network during acupuncture stimulation
reveals a striking similarity to the default mode network during
attention-demanding tasks (Hui et al., 2009). The task-positive
network that is anticorrelated to the default mode network in the
resting brain shares a few common regions with the activation net-
work in acupuncture, such as the sensorimotor cortex and paralimbic
In this review we summarize our research, discuss the brain
networks involved in acupuncture, and describe their relationships.
We then discuss future research directions, including the role of
neurotransmitters in acupuncture action, the difference between
acupuncture methods, and the selection of and differences among
acupuncture control stimulation. We then conclude with a summary
of the potential clinical application of our research ﬁndings given
what is currently known about several common disease states that
affect the brain.
2. Our research
Over the past decade, we have built a database of fMRI scans of the
brain response to traditional Chinese acupuncture at multiple
acupoints in healthy adults. We have focused on three classical
acupoints that are commonly used clinically for their analgesic and
regulatory clinical effects, namely LI4 (hegu) on the hand, ST36
(zusanli) on the leg, and LV3 (taichong) on the foot (Hui et al., 2000,
2005, 2009). In our studies we have compared four minutes of
acupuncture needling to ﬁve minutes of the acupuncture needle at
rest, in a standard ten-minute paradigm, excluding the ﬁrst minute of
scanning during the ﬁrst rest period to allow for baseline equilibration
of the fMRI machine, (Fig. 1).
We ﬁrst described the coordinated signal decreases in the limbic
system in fMRI of the brain during acupuncture at LI4 in Hui et al.
(2000). This partial brain imaging encompassed cortico–limbic and
sensorimotor regions of special interest to the study. Our observation
of decre ased signal represented decreased blood ﬂow during
acupuncture needling in brain areas including the nucleus accumbens,
amygdala, hippocampus, parahippocampus, hypothalamus, ventral
tegmental area, anterior cingulate gyrus, caudate, putamen, temporal
pole, and insula. These decreases were found in patients experiencing
the constellation of sensations termed deqi, but were absent or
markedly attenuated in patients experiencing sharp pain (Figs. 2
and 3). In contrast, signal increases in the somatosensory cortices
were present during both acupuncture and sensory control. In this
early paper we ﬁrst showed evidence for the coordinated signal
decreases during BOLD fMRI, representing decreased cerebral blood
ﬂow due to suppressed neuronal metabolic activity within the limbic
and paralimbic systems.
In Hui et al. (2005) we extended our fMRI ﬁndings to ST36,
showing with whole brain imaging that manual acupuncture at this
point also led to coordinated decreased signal throughout the limbic
system. We also described the central effect of acupuncture on the
cerebellum, showing that the cerebro–cerebellar system also dem-
onstrated signal decreases in concert with the limbic system. Again,
the pattern of hemodynamic response depended on the psychophys-
ical response to needle manipulation, with
deqi sensations leading to
signal decrease, while inadvertent sharp pain led to signal increase.
Tactile stimulation as control elicited signal increases, predominantly
in the somatosensory areas. Based on these ﬁndings, we argued that
the cerebro–cerebellar and limbic system responds in an integrated
manner in correlation with the psychophysical response to acupunc-
ture stimulation at ST36.
In our most recent study (Hui et al., 2009), we described our
analysis of 201 acupuncture runs and 74 tactile stimulation control
runs in 48 healthy subjects at all three acupuncture points. The large
size of this data set provides a foundation for compariso n of
neuroimaging ﬁndings in health and disease and between acupunc-
ture and tactile stimulation controls. We again described clusters of
decreased activity in limbic and paralimbic regions including the
medial prefrontal, medial parietal and medial temporal lobes, along
with increased activit y in the sensorimotor cortices and select
paralimbic structures. To better analyze and characterize this large
fMRI dataset, we used a general linear model and cross-correlation
analysis to identify the activation and deactivation networks and their
functional connectivity during acupuncture administration (Fig. 4).
The temporal and spatial features of the activation and deactiva-
tion networks were compared with descriptions in the resting brain
literature to explore their relationships as well as additional effects
that acupuncture might evoke. We found that the extensive regions of
deactivation and activation observed during acupuncture showed
substantial overlap with both the default mode network and the anti-
correlated task-positive network in response to stimulation (Fig. 5).
However, during acupuncture, the amygdala and hypothalamus—
structures not commonly reported as involved in the default mode
literature—were found to be a central part of the activation response
of the brain to acupuncture.
A second reproducible ﬁnding of our research has been that the
signal increase observed in the somatosensory regions during acupunc-
ture needling, an invasive procedure, is less than the signal increase seen
during the control superﬁcial tactile stimulation. Tactile stimulation, as
has been previously described, induced greater activation in the
somatosensory regions but less extensive involvement of the default
mode network and limbic-associated regions of the anti-correlated
task-positive network. Moreover, even short periods of inadvertent pain
can attenuate or reverse the deactivation of the default mode network
that occurs during acupuncture. Such effects indicate that deactivation
of the default mode network during acupuncture cannot be completely
explained by the demand of attention, as is commonly proposed in the
literature. Together, these results strongly suggest that acupuncture
engages extensive functionally organized intrinsic systems of the brain
as mediators of its diverse effects.
To provide a single terminology for the network of brain regions
involved in the acupuncture response, we have deﬁned the limbic–
paralimbic –neocortical network, comprised of the default mode
network, amygdala and hypothalamus (Figs. 6 and 7). Our ﬁndings
show that acupuncture mobilizes these anti-correlated functional
networks of the brain to mediate its actions, and that the effect is
dependent on the psychophysical response to acupuncture stimula-
tion. Functional MRI studies from other groups have also shown
generalized deactivation of the limbic–paralimbic–neocortical net-
work across multiple levels of the brain, along with activation of the
sensorimotor system, during acupuncture stimulation (Fang et al.,
Fig. 1. Time course of acupuncture needling. The acupuncture needle was inserted and
the sensitivity of the subject to manipulation was pre-tested and adjusted to tolerance
prior to starting functional MRI scanning. The needle remained at rest for 2 min after
the start of MRI scanning before bidirectional rotation at 1 Hz for 2 min. The needle was
not manipulated for 3 min, then manipulation was repeated for 2 min, followed by a
third period of rest for 1 min. The needle was removed after MRI scanning was
complete. Data analysis compared the blood oxygenation level-dependent (BOLD) MRI
signal intensity of the two needling periods with the three rest periods, with the ﬁrst
minute of scanning excluded from analysis to allow for MRI signal equilibration.
82 K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
2006; Fang et al., 2008; Napadow et al., 2005a; Wang et al., 2007; Wu
et al., 1999). The relevance of our ﬁ ndings to the clinical effect of
acupuncture is supported by a recent study showing deactivation of
the default mode network and activation of the somatosensory cortex
persisting for more than 10 min after the completion of electro-
acupuncture stimulation (Dhond et al., 2008; Bai et al., 2009).
Fig. 2. The inﬂuence of sensations on brain fMRI signal changes during acupuncture and sensory control at ST36. Group average functional results showing signal decreases (blue)
and increases (red), thresholded at p b 0.001 and a 3-voxel cluster size. All slices are 2 mm parasagittal in the right hemisphere in Talairach space (Talairach and Tournoux, 1988).
(Left) Acupuncture with deqi sensations without sharp pain (n =11) resulted in widespread signal decreases. (Center) Acupuncture with mixed deqi and sharp pain sensations
(n= 4) resulted primarily in signal increases. (Right) Sensory control (n= 5) also resulted in signal increases beyond dedicated sensorimotor areas. Regions: (1) frontal pole,
(2) subgenual cingulate, (3) ventromedial prefrontal cortex, (4) hypothalamus, (5) posterior cingulate, (6) reticular formation, (7) cerebellar vermis, (8) middle cingulate, and
Fig. 3. Brain activity with acupuncture at ST36 in a single subject who reported deqi sensations without sharp pain during one experimental run and only sharp pain without
additional sensations during a subsequent experimental run. The broad degree of deactivation (blue) during acupuncture with deqi is in contrast to the general activation (red) noted
during acupuncture with sharp pain alone. Time-course for the voxel with peak signal change within the right amygdala is shown for each run. Regions: (1) pregenual cingulate/
frontal pole, (2) posterior cingulate Brodmann area 31/precuneus, (3) substantia nigra, (4) middle cingulate, (5) thalamus, (6) periaqueductal gray, (7) cerebellar tonsil,
(8) amygdala, (9) parahippocampus, (10) insula.
83K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
Fig. 4. Seed-based cross-correlation analysis of deactivation network in 48 matched subjects in 201 acupuncture runs with deqi and 74 tactile stimulation runs with and without deqi.
Reference regions are circled in red, and regions with positive correlation activity are circled in green. Correlations with pb 0.001 are shown. (A, B) Correlations for the posterior
cingulate Brodmann area 31: (A) Acupuncture: reference voxel (2, − 53, 36), showing correlations in the medial prefrontal, posterior medial parietal, medial temporal lobe
and temporal po le. (B) Tactile st imulation: reference voxel (3, − 63, 31), showing correlations that partially overlap those in acupuncture, but correlations are markedly weaker.
(C,D) Correlations for the hypothalamus: (C) Ac upuncture: reference voxel (3, − 1, − 10) , showing correlations with regi ons similar to (A), but more limited in extent. (D ) Tactile
stimulation: reference voxel (3, − 1, − 10), showing no regions with signiﬁcant signal change. Numbered ar eas: (1) frontal pole and pregenual ci ngulate, (2) subge nual cingulate,
and subgenual area 25 in the ventral route of the medial prefrontal cortex, (3) precuneus, posterior cingulate and retrosplenial cortex of the posterior medial parietal cortex,
(4) amygdala, hippocampus, and par ahippocampus of the medial temporal lobe, (5) temporal pole, (6) precuneus and posterior cingulate cortex Brodmann area 31, and
(7) orbitofrontal cortex, subgenual cingulate, subgenual area SG25 and ventromedial prefrontal c ortex.
Fig. 5. Comparison of the limbic–paralimbic–neocortical network (LPNN) during acupuncture deqi and the default mode network (DMN) as described in the literature (adapted from
Shulman et al., 1997). Multiple sagittal slices through an averaged brain in Talairach space (Talairach and Tournoux, 1988) are shown for the LPNN, spanning both the right
hemisphere (left) and the left hemisphere (right), with distance from midline shown in mm. Two brain surface renderings (Cox, 1996) are also shown in the lower left. An oblique
view of the DMN is shown in the lower right.
84 K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
2.1. Sensations and deqi
Throughout our studies, subjects have rated the sensations they
experienced during acupuncture stimulation, and these sensation
ratings have been central to our analyses. Subjects rated 10 sen-
sations, namely, a ching, soreness, pressure, f ullness, heaviness,
numbness, tingling, warmth, coolness, and dull pain, on a scale from
0 (none) to 10 (unbearably strong). According to traditional Chinese
Medicine teaching, the clinical efﬁcacy of acupuncture is related to the
unique psychophysical response of deqi, which consists of these
sensations. Subjects also were asked to report during the acupuncture
stimulating runs the occurrence of sharp pain. When sharp pain was
reported, the therapist reduced the rate of stimulation until the pain
subsided, which took place within seconds. We characterized and
detailed subjects' experience of these sensations during acupuncture
in Hui et al. (2007). The characterization and deﬁnition of deqi is
critical to discriminating between the effects of acupuncture stimu-
lation, as we have repeatedly found that the hemodynamic response
to acupuncture is dependent on the psychosomatic response. This is
further supported by the link of the limbic system to both emotions
Throughout our studies we have found that 71% of acupuncture
and 24% of tactile stimulation control runs elicited deqi, using the sum
score of 3 across the 10 tested sensations as a minimal cutoff value for
deﬁning deqi. Patient sensations can be categorized as deqi, mixed deqi
with sharp pain, or, very rarely, sharp pain alone. The sensations most
associated with deqi were aching, soreness and pressure; these
ﬁndings were consistent with the literature. Differences were found
among acupuncture points, with LI4 showing the most prominent
response. The data provides scientiﬁc support for the general belief
that LI4 is the most potent acupuncture point in traditional Chinese
acupuncture, and is therefore most often employed in clinical practice
(Napadow et al., 2004).
Moreover, we have noted a novel relationship between dull pain
and sharp pain. In neurophysiology, dull pain is commonly described
as a slow or secondary pain that follows sharp pain (Snell, 2006). The
order is reversed in our acupuncture studies, with dull pain often
occurring independently of sharp pain and, at times, as incipient sharp
pain. The reason for the discrepancy between our ﬁndings on pain and
those in the literature is unclear; it may be related to the nature of the
stimulus. While most acute pain studies employ heat or a sharp device
that targets the skin surface, the gentle motion of a smooth ﬁne needle
in deeper tissue layers may elicit a more diffuse stimulation of the A
delta or c ﬁbers, without reaching the threshold of overt noxious
simulation (Figs. 2 and 3).
2.2. Deactivated brain networks in acupuncture
The brain regions that preferentially show a coherent decrease in
activity during acupuncture constitute the limbic–paralimbic–neo-
cortical network. Coherent deactivation of a subset of these task-
negative regions was also noted, to a lesser degree, during tactile
stimulation. This signiﬁcantly weaker change in signal intensity of
tactile stimulation was also more limited in spatial distribution. The
difference between the brain response to acupuncture and tactile
stimulation was especially marked in the pregenual cingulate,
posterior cingulate, and precuneus on both hemispheres, suggesting
the presence of subsystems that differ in their sensitivity to the two
types of sensory stimuli. Statistical signiﬁcance was reached with our
large sample size.
The deactivation of the amygdala and hypothalamus during
acupuncture but not during tactile stimulation suggests a speciﬁc
role for these major limbic structures in acupuncture action. These
limbic structures are central to the regulation and control of emotion,
cognition, bio-behavior, endocrine and autonomic nervous functions,
and are activated by stress, pain and emotions of negative valence.
Major depression and other mood disorders are characterized by
amygdala hypers ensitivity to activation by negative emotional
stimuli. To our knowledge, the amygdala was included in only two
reports (Lowe et al., 1998; Shulman et al., 1997) and the hypothal-
amus in only one report (Greicius et al., 2003) among the many
Fig. 6. Limbic–Paralimbic–Neocortical Network activity during acupuncture deqi.The
averagedBOLD signal changes are shownfor 11 subjects during acupuncture atpoint ST36,
showing activation (red) and deactivation (blue). A parasagittal image is shown at 2 mm
from midline within the right hemisphere in Talairach space (Talairach and Tournoux,
1988). Regions: (1) frontal pole, (2) ventromedial prefrontal cortex, (3) pregenual
cingulate, (4) subgenual cingulate, (5) subgenual Brodmann area 25, (6) septal nuclei,
(7) hypothalamus, (8) posterior cingulate, 9 precuneus, (10) substantia nigra, (11) retic-
ular formation, and (12) cerebellar vermis.
Fig. 7. Functionally anti-correlated task-positive (activation, red, italic font) and
negative (deactivation, blue, regular font) networks involved in acupuncture action
shown on a brain sketch, with an overlapping of medial and lateral regions. Regions of
deactivation in the limbic–paralimbic–neocortical network aggregate in the medial
frontal, parietal and temporal lobes. The basal ganglia, the cerebellar vermis, tonsil and
brainstem also show regional deactivation. The cingulate and thalamus have regions in
both networks, while the sensorimotor areas, insula, middle cingulate and the dorsal
division of posterior cingulate Brodmann area 23 are in the task-positive network. The
ﬁgure is adapted from Hui et al. (2009). Abbreviations: ant middle C, anterior middle
cingulate; BA Brodmann area; BA23d, Brodmann area 23 dorsal; BA23v, Brodmann area
23 ventral; BG, basal ganglia; FP, frontal pole; Hyp, hypothalamus; M1/S1, primary
motor/primary sensory cortex; OFC, orbitofrontal cortex; PAG, periaqueductal gray; PB,
parabrachial nucleus; preg C, pregenual cingulate; SII, secondary somatosensory cortex;
SMA, supplementary motor area; SG25, subgenual area 25; Subg C, subgenual cingulate,
TA, tract A; TB, tract B.
85K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
publications on the response of the default mode network to cognitive
tasks without emotional components.
Other groups have reported the effects of acupuncture on the
default mode network during the resting state more than 10 min after
acupuncture (Bai et al., 2009; Dhond et al., 2008). A substantial
overlap was found with the response pattern described for the default
mode network in the literature. The amygdala showed deactivation in
both studies in agreement with our results, and the DMN literature.
The hypothalamus showed activation in Bai's study, as opposed to the
deactivation seen in our data base and in a DMN report (Greicius et al.,
The coherent deactivation of the temporal pole described in ourmost
recent study (Hui et al., 2009) also appears to be more extensive than in
the current literature on the default mode network. The region has close
anatomical interconnections with other medial temporal lobe and
prefrontal cortical regions, and is thought to play an important role in
memory and social emotional processing (Kahn et al., 2008; Olson et al.,
2007). Taken together, the involvementof the amygdala, hypothalamus,
and temporal pole in acupuncture may represent a task-negative
network subsystem, either within or outside the default mode network,
which is mobilized by acupuncture.
The functional connectivity of the default mode network and the
additional limbic and paralimbic regions observed during acupunc-
ture is supported by anatomical connectivity demonstrated by
multiple approaches (Buckner et al., 2008; Honey et al., 2009). A
recent study provided a large scale map of the interconnections
between the core components of the default mode network in the
macaque (Parvizi et al., 2006), adding to prior anterograde and
retrograde tracer studies in animals, including primates (Schmah-
mann and Pandya, 2006). A dditionally, modern techni ques in
diffusion tensor and diffusion spectral imaging have enabled the
demonstration in vivo of the major white ﬁber paths of these circuits
in human subjects (Fujiwara et al., 2008; Hagmann et al., 2007; Honey
et al., 2009; Wedeen et al., 2008).
2.3. Activated brain networks during acupuncture
The activation of sensor imotor and heteromodal association
cortices, the secondary somatosensory cortex, the supplementary
motor area, the dorsolateral prefrontal cortex, the right anterior
insula, the anterior middle cingulate and the dorsal divisions of the
posterior cingulate Brodmann area 23 observed during acupuncture
(Hui et al., 2009) overlaps with that activation of the task-positive
network described for the resting brain (Fox et al., 2005; Fransson,
2005). The task-positive system includes the dorsal attention system
(Corbetta and Shulman, 2002) and the executive control system
(Seeley et al., 2007). Contrary to conventional expectation, our
functional MRI studies indicate that activation of the sensorimotor
cortex by invasive acupuncture needling is weaker than the activation
elicited by gentle tapping with a ﬂexible nylon monoﬁlament (Hui et
al., 2000, 2005, 2009). Sensorimotor cortex activation by acupuncture
becomes stronger than tactile control only when acupuncture
provokes sharp pain. We hypothesize that such a dichotomy may be
due to the different pathways taken by different neural impulses
arising from peripheral sensory receptors (see tracts A and B in Fig. 7).
Activation of the sensorimotor cortex is primarily mediated by
impulses ascending via the dorsal column medial lemniscal system,
while modulation of the limbic system is primarily mediated by
impulses ascending via the spinocervical, spinoreticular and spino-
mesencephalic tracts (Willis and Westlund, 1997). Studies of animal
models demonstrate that axons in these latter ﬁber tracts send
collaterals to synapse directly with neurons in the dorsal thalamus,
dorsal midbrain, medullary and pontine reticular formation, hypo-
thalamus, amygdala, septum and nucleus accumbens of the limbic
system (Willis, 1989; Willis and Westlund, 1997). We postulate that a
greater proportion of impulses generated by acupuncture stimulation
results in signals that reach the limbic system to exert its modulatory
effects, whereas a greater proportion of impulses generated by tactile
stimulation results in signals that reach the sensorimotor cortex
to exert its excitatory effects. Such differentiation by pathway
would explain the overlap of activation and deactivation, as well as
the differences in the predominance of effects between these two
2.4. Relationship between deactivation and activation networks
It is suggested that the anti-correlation between the default mode
network and the task-positive networks may reﬂect the dichotomy
between increased brain activity in regions supporting execution of a
task and decreased brain activity in regions involved in unrelated
processes (Fox et al., 2005). The partition of paralimbic structures
such as the cingulate into both the anti-correlated deactivation and
activation networks can be explained by their differences in
cytoarchitecture and function (Vogt, 2005; Vogt et al., 2003, 2006).
For example, the middle cingulate, which is involved in attention and
pain perception, is activated, whereas both the rostral divisions of the
cingulate, which are involved in emotion, autonomic and salience
processing, and the caudal divisions of the cingulate, which are
involved in self-reference assessment, are deactivated.
The presence of subsystems has been proposed for the anti-
correlated networks of the resting brain (Golland et al., 2008).
Comparisons of acupuncture with conventional tactile stimulation
and noxious stimulation suggests that there may be multiple variables
and sub-systems involved. A subset of regions within the task-positive
network, including the right anterior insula, thalamus, anterior
middle cingulate and dorsal division of the posterior cingulate
Brodmann area 23 (all limbic-related structures), appear to be more
strongly conne cted and anti-corr elated with the task-negative
network than the subset of regions comprising sensorimotor and
association cortices such as the second somatosensory cortex and the
dorsolateral prefrontal cortex. These limbic-related task-positive
structures show strong responses to acupuncture, with and without
noxious pain, but markedly less response during tactile stimulation.
Meanwhile, the sensorimotor and hetereomodal association cortices
may form another sub-system that responds more to tactile
stimulation and noxious pain than to acupuncture (see Fig. 7).
3. Future directions and experimental considerations
3.1. Involvement of neurotransmitters: beyond endorphins
Although brain function can be modeled and discussed based on
brain regions and interconnections, the underlying mechanism is also
based on a complex interplay of neurotransmitters and neuromodu-
lators. The endogenous opioid peptides and serotonin are known for
their role in analgesia (Han, 2004), but may not account fully for the
affective dimension of pain and the diverse modulatory effects of
acupuncture. Instead, excitatory and inhibitory mediators of smaller
size may play a central role, at least in the early phase.
Of the monoamines, dopamine is the neurotransmitter of highest
concentration in the limbic system (Cooper et al., 2002). Animal data
acquired by different experimental approaches indicate that acu-
puncture suppresses the synthesis or release of dopamine induced by
pain (Shi et al., 1986), cocaine (Yang et al., 2001) or alcohol (Yoon et
al., 2004). Our study in rats demonstrated that electroacupuncture on
the forepaw reversed activation of the limbic system induced by
amphetamine, accompanied by a marked rise of GABA and return of
elevated glutamate and dopamine levels to baseline levels in striatal
microdialysates (Chen et al., 2008). The signal increase of the median
raphe nuclear groups observed in the present study as well as in an
earlier study (Napadow et al., 2005a) is consistent with serotonergic
activation. The deactivation of the subgenual cingulate and the
86 K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
reticular formation (Hui et al., 2005, 2009) is consistent with reports
of sympathetic tone down-regulation by acupuncture (Cao et al.,
1983; Haker and Bjerring, 2000). Thus, the patterns of hemodynamic
response observed in fMRI are in accord with the current knowledge
of acupuncture effects on the monoaminergic mediators and auto-
nomic system function.
The important role of GABAergic inhibition in the maintenance of
brain function is receiving increasing attention. Many studies point to
a relation between a decrease of GABAergic tone and interneuron
dysfunction in the cortico-limbic system in schizophrenia, depression
and anxiety disorders (Benes, 2009). Another amino acid, glycine, is
being explored for its inhibitory action (Delaney et al., 2009). The
neuronal communication of neural circuits is complex and must be
viewed in an integrative way. Much more work is needed to deﬁne the
components that play major roles at different stages and at different
levels of the cortico-limbic system during acupuncture treatment. The
complex interplay in the neural circuits is unlikely to be adequately
explained by one or two transmitters or modulators or by a few brain
3.2. Acupuncture methods
In practice, acupuncture is performed in a variety of ways.
Traditional acupuncture is performed with needle manipulation,
and is most commonly used in clinical practice (Napadow et al., 2004).
Most practitioners employ gentle stimulation in routine settings,
aiming to generate deqi without provoking sharp pain. However, the
technique and intensity of manipulation varies by individual, ranging
from little or no manipulation to rapid, painful needling. Modern
methods of acupuncture include electroacupuncture, where a small
electric current is applied to needles inserted into acupoints, or
transcutaneous electric neural stimulation, where a small electric
current is applied through electrodes placed on the skin. The extent to
which different methodology alters the effects of acupuncture in the
clinical domain remains to be studied.
In vivo studies of the mechanism of acupuncture action in humans
was ﬁrst enabled by the advent of functional MRI. For our human
subjects research program we have rigorously controlled the
technique of acupuncture delivery. We used a block design (Fig. 1)
with gentle bidirectional needle rotation at 1 Hz for the acupuncture
stimulation periods. The needling frequency was reduced when
subjects signaled very strong sensations or sharp pain during needling
by raising one or two ﬁngers, respectively. As a result, only rarely did a
subject have more than a brief episode of sharp pain during our
experiments. Moreover, in our data analysis, we separated subjects
who experienced sharp pain from the major cohort who experienced
deqi without sharp pain to avoid the confounding effect of the noxious
stimulus. To our knowledge, most studies currently in the literature
fail to differentiate subjects in this way. A confounding effect may be
particularly marked in limbic regions associated with emotional
processing and the affective dimensions of pain. These or other
differences in methodology may contribute to discrepancies in the
hemodynamic response in the acupuncture neuroimaging literature,
for instance, the report activation versus deactivation of cortico-limbic
structures (Bai et al., 2009; Wu et al., 1999; Liu et al., 2007).
Stimulation by other modalities, such as electro-acupuncture,
often at different frequencies, adds another dimension of complexity
to the comparison. Variation in the effect of different frequencies of
electroacupuncture has been previously reported in a study of
transcutaneous electrical nerve stimulation (TENS) for primary
dysmenorrhea (Lundeberg et al., 1985), where high-frequency TENS
was found more effective than low-frequency TENS at reducing
symptoms, and where the effect of low- but not high-frequency TENS
was found to be reversed by naloxone, a relatively pure opiate
antagonist. An imaging study from our group (Napadow et al., 2005b)
also showed a difference, with low frequency electro-acupuncture at
ST36 found to be generally similar to manual acupuncture in down-
regulating corticolimbic activity as determined by functional MRI,
while high-frequency electroacupuncture was not.
Finally, the site of acupuncture stimulation may also be a factor.
We have chosen acupuncture points located in muscle or in tendino-
muscular layers that are widely used for their a nalgesic and
modulatory effects. Acupuncture points with different histological
and nerve supply characteristics remain to be explored. Multiple lines
of research must be pursued in order to characterize the full spectra of
stimulations that are commonly referred to as acupuncture.
3.3. Experimental design — appropriate controls for acupuncture studies
The choice of an appropriate control condition for acupuncture
research is highly controversial. Based on the ubiquity of sensory
receptors and nerve ﬁbers, sensory stimulation generates impulses
that target the brain, regardless of the site of stimulation or degree of
invasiveness. Multiple approaches to control stimulation for acu-
puncture have been suggested. These include acupuncture needling at
points not located on acupuncture meridians, minimized acupuncture
stimulation with the needle inserted into an acupuncture point only
superﬁcially, or superﬁcial tactile stimulation over an acupuncture
point. Although the pattern and degree of response to these
stimulations may vary, no skin location can be considered inert
(Lund and Lundeberg, 2006).
We opted to use tactile stimulation of the same acupuncture points
as were subsequently stimulated using traditional acupuncture by
tapping gently with a von Frey monoﬁlament of standard force.
Although in some experimental paradigms the subjects may visually
observe the difference in stimulation method, in our studies subjects
were in the MRI machine during stimulation and could not observe
this difference. Consistent results have been observed for multiple
acupoints, and administered by different acupuncturists in our studies
over the years.
We believe that acupuncture performed at the 3 acupoints used in
our studies primarily stimulates the receptors and nerve ﬁbers in the
deeper tendino-muscular tissue layers, while tactile stimulation
primarily stimulates those in the skin. Thus, our data suggests that
stimulation of the tendino-muscular tissue may predomina ntly
account for the extensive effects on the limbic–paralimbic–neocorti-
cal network that we observed acutely during acupuncture, and that
stimulation of the sensory elements on the skin may predominantly
account for the stimulatory ef fect on the sensorimotor cortical
regions. One possible mechanism of the central effect of acupuncture
may be through the low-threshold mechanoreceptors found in hairy
skin which are innervated by unmyelinated C tactile afferents. These
activate the insular cortex in a manner well-suited to encoding slow,
gentle touch, but only poorly encoding discriminative aspects
(Olausson et a l., 2010). Whether the mild deactivation of the
limbic–paralimbic–neocortical network observed during tactile stim-
ulation reﬂects a mild stimulation of the same pathway as
acupuncture action, or whether tactile stimulation simply represents
a salient stimulus that competes with the intrinsic networks of the
brain, leading to apparent deactivation during the task, remains an
open and important question.
3.4. Potential clinical beneﬁts of acupuncture therapy
Acupuncture has been shown to be promising in the treatment of
several disease states. A study from our group on longstanding carpal
tunnel syndrome suggests that acupuncture promotes the recovery of
the alterations in synaptic plasticity found in this disorder (Napadow
et al., 2007). Preliminary results of randomized controlled trials
suggest that acupuncture may be effective in the treatment of major
depression ( Leo and Ligot, 2007; Roschke et al., 2000; Wang et al.,
2008) and post -traumatic stress disorder (Holliﬁeld et al., 2007). The
87K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
potential for the application of acupuncture is found in the extensive
literature on alterations in the functioning of the default mode
network as described in multiple disease states, including Alzheimer's
disease, with impaired connectivity between the medial prefrontal
lobe and the posterior medial parietal cortex (Greicius and Menon,
2004; He et al., 2007; Lustig et al., 2003; Rombouts et al., 2005; Wang
et al., 2006); autism, with almost complete loss of connectivity
between major core regions (Kennedy and Courchesne, 2008;
Kennedy et al., 2006); chronic low back pain, with atrophy of the
prefrontal cortex (Baliki et al., 2008); schizophrenia (Bluhm et al.,
2007; Liang et al., 2006; Zhou et al., 2007); major depression (Anand
et al., 2007; Drevets, 2007; Greicius et al., 2007); attention deﬁcit
hyperactivity disorder (Cao et al., 2006; Castellanos et al., 2008; Zang
et al., 2007); multiple sclerosis (Lowe et al., 2002); and Parkinson's
disease (Stoffers et al., 2007 ). Additional therapeutic effects for
acupuncture have been shown for disorders involving affective states
such as anxiety, depression and substance abuse (Edzard et al., 1998;
Margolin et al., 1993), where the acupuncture effects are similar to
those of deep brain stimulation (Mayberg et al., 2005; Lozano et al.,
Interestingly, two psychostimulant drugs, cocaine and nicotine,
induced bilateral fMRI signal increases in many brain regions ( Breiter
et al., 1997; Stein et al., 1998), a subset of which demonstrated signal
decreases during acupuncture needle manipulation. The opposite
effects of acupuncture needle manipulation and psychostimulants on
these limbic and paralimbic regions suggest that acupuncture
treatments could be effective for drug craving and detoxiﬁcation.
This is supported by a meta-analysis of trials of opioid receptor
agonists with and without acupuncture for the treatment of opioid
detoxiﬁcation (Liu et al., 2009), which argues that acupuncture use
decreases side effects and the dose of opioid agonist needed during
detoxiﬁcation. We can further speculate that acupuncture analgesia
for surgical procedures may work by decreasing neural activity in the
thalamus, amygdala, and brainstem, structures that are known to
modulate the conscious experience of pain (Becerra et al., 1999; Casey
et al., 1994; Coghill et al., 1994; Craig et al., 1996; Davis et al., 1997;
Jones et al., 1991; Talbot et al., 1991). Future research on the effect of
acupuncture on these and other disorders is necessary to evaluate
whether our ﬁndings in healthy subjects translates to the treatment of
3.5. Acupuncture and traditional Chinese medicine
Acupuncture is deeply rooted in the yin-yang theory of traditional
Chinese medicine. The Chinese refer to yin and yang as the natural
polarity in all things in the universe, for example dark and light, heat
and cold, negative and positive, or inhibitory and excitatory.
Traditional Chinese medicine holds that that acupuncture heals by
reestablishing the balance between yin and yang and restoring the
normal ﬂow of vital energy, called qi, which has become impaired in
disease states. Although the anatomic and physiologic substrate of qi
is not completely clear, clinical and experimental data indicate that
many of the effects are mediated via identiﬁable components of the
nervous system (Lin and Wang, 1994; Cheng, 2000). With the
demonstration of the naturally occurring anti-correlated systems in
the brain that interact to maintain normal functions and a state of
health, it may be possible to draw analogies between yin and the
deactivation of the task-negative default mode network, yang and the
activation of the task-positive network, and the ﬂow of qi and
functional connectivity. This is speculative, but is nonetheless an
intriguing relationship that may guide future investigations.
We have explored the effect of acupuncture on the brain in healthy
subjects through several studies. Together these studies show that the
sensations constituting deqi are associated with decreased brain
activity in the limbic system and in the default mode network, while
sharp pain is generally associated with signal increases in these same
regions. To better group the set of regions involved in the response to
acupuncture, we have deﬁned the limbic–paralimbic–neocortical
network, consisting of the amygdala, hypothalamus, and default
mode network, thereby encompassing the limbic system. At the same
time, we have shown that tactile sensory stimulation as control
primarily affects somatosensory areas. Given our ﬁndings, we propose
that the observed effect on the limbic system by acupuncture is
central to its mechanism of action, and goes beyond the effect of
attention on the default mode network, as described in the literature.
There are multiple avenues of future research to be pursued. To
what extent neurotr ansmitters play a role in the response to
acupuncture is little understood, and may be elucidated through
animal models or by developing avenues such as magnetic resonance
spectroscopy or positron emission tomography. The extent to which
differences in techniques affect the central effects and its clinical
relevance needs further exploration. Finally, the clinical use of
acupuncture needs further testing through well-designed clinical
trials. We believe our ﬁndings of the central effect of acupuncture on
the limbic system supports the testing of acupuncture for therapeutic
use in affective and cognitive disease states, as well as its traditional
role in the treatment of chronic pain.
The work was supported in part by the NIH/National Center for
Complementary and Alternative Medicine (R21-AT00978) (1-P01-
AT002048-01) ( 2-P01-AT002048-06) (K01-AT-002166-01), (F05-
AT003770), the National Center for Research Resources (P41RR14075),
the Mental Illness and Neuroscience Discovery Institute and the Brain
Project Grant NS 34189. We wish to thank Nikos Makris for consultations
on neuroanatomy, and Randy Buckner for useful discussions on the
default mode system. There are no conﬂicts of interest for any author.
Anand, A., Li, Y., Wang, Y., Gardner, K., Lowe, M.J., 2007. Reciprocal effects of
antidepressant treatment on activity and connectivity of the mood regulating
circuit: an FMRI study. J. Neuropsychiatry Clin. Neurosci. 19, 274–282.
Bai, L., Qin, W., Tian, J., Dong, M., Pan, X., Chen, P., Dai, J., Yang, W., Liu, Y., 2009.
Acupuncture modulates spontaneous activities in the anti-correlated resting brain
networks. Brain Res. 1279, 37–49.
Baliki, M.N., Geha, P.Y., Apkarian, A.V., Chialvo, D.R., 2008. Beyond feeling: chronic pain
hurts the brain, disrupting the default-mode network dynamics. J. Neurosci. 28,
Becerra, L.R., Breiter, H.C., Stojanovic, M., Fishman, S., Edwards, A., Comite, A., Gonzalez,
R.G., Borsook, D., 1999. Human brain activation under controlled thermal
stimulation and habituation to noxious heat: an fMRI study. Magn. Reson. Med.
Benes, F.M., 2009. Neural circuitry models of schizophrenia: is it dopamine, GABA,
glutamate, or something else? Biol. Psychiatry 65, 1003–1005.
Binder, J.R., Frost, J.A., Hammeke, T.A., Bellgowan, P.S., Rao, S.M., Cox, R.W., 1999.
Conceptual processing during the conscious resting state. A functional MRI study.
J Cogn. Neurosci. 11, 80–95.
Bluhm, R.L., Miller, J., Lanius, R.A., Osuch, E.A., Boksman, K., Neufeld, R.W., Theberge, J.,
Schaefer, B., Williamson, P., 2007. Spontaneous low-frequency ﬂuctuations in the
BOLD signal in schizophrenic patients: anomalies in the default network.
Schizophr. Bull. 33, 1004–1012.
Breiter, H.C., Gollub, R.L., Weisskoff, R.M., Kennedy, D.N., Makris, N., Berke, J.D.,
Goodman, J.M., Kantor, H.L., Gastfriend, D.R., Riorden, J.P., Mathew, R.T., Rosen, B.R.,
Hyman, S.E., 1997. Acute effects of cocaine on human brain activity and emotion.
Neuron 19, 591–611.
Buckner, R.L., Andrews-Hanna, J.R., Schachter, D.L., 2008. The brain's default network:
anatomy, function, and relevance to disease. Ann. NY Acad. Sci. 1124, 1–38.
Cao, X.D., Xu, S.F., Lu, W.X., 1983. Inhibition of sympathetic nervous system by
acupuncture. Acupunct. Electrother. Res. 8, 25–35.
Cao, Q., Zang, Y., Sun, L., Sui, M., Long, X., Zou, Q., Wang, Y., 2006. Abnormal neural
activity in children with attention deﬁcit hyperactivity disorder: a resting-state
functional magnetic resonance imaging study. Neuroreport 17, 1033–1036.
Casey, K.L., Minoshima, S., Berger, K.L., Koeppe, R.A., Morrow, T.J., Frey, K.A., 1994.
Positron emission tomographic analysis of cerebral structures activated speciﬁcally
by repetitive noxious heat stimuli. J. Neurophysiol. 71, 802–807.
88 K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
Castellanos, F.X., Margulies, D.S., Kelly, C., Uddin, L.Q., Ghaffari, M., Kirsch, A., Shaw, D.,
Shehzad, Z., Di Martino, A., Biswal, B., Sonuga-Barke, E.J., Rotrosen, J., Adler, L.A.,
Milham, M.P., 2008. Cingulate–precuneus interactions: a new locus of dysfunction
in adult attention-deﬁcit/hyperactivity disorder. Biol. Psychiatry 63, 332–337.
Chen, Y.I., Ren, J.Q., Wang, F.N., Xu, H.B., Mandeville, J.B., Kim, Y., Rosen, B.R., Jenkins, B.G.,
Hui, K.K.S., Kwong, K.K., 2008. Inhibition of stimulated dopamine release and
hemodynamic response in the brain through electrical stimulation of rat forepaw.
Neurosci. Lett. 431, 231–235.
Cheng, X.N., 2000. Chinese Acupuncture and Moxibustion. People's Medical Publishing
House, Beijing. 590 pp.
Coghill, R.C., Talbot, J.D., Evans, A.C., Meyer, E., Gjedde, A., Bushnell, M.C., Duncan, G.H.,
1994. Distributed processing of pain and vibration by the human brain. J. Neurosci.
Cooper, T.R., Bloom, F.E., Roth, R.H., 2002. The Biochemical Basis of Neuropharmacology.
Oxford Academic Press, New York. 528 pp.
Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and stimulus-driven
attention in the brain. Nat. Rev. Neurosci. 3, 201–215.
Cox, R.W., 1996. AFNI: software for analysis and visualization of functional magnetic
resonance neuroimages. Comput. Biomed. Res. 29, 162–173.
Craig, A.D., Reiman, E.M., Evans, A., Bushnell, M.C., 1996. Functional imaging of an
illusion of pain. Nature 384, 258–260.
Davis, K.D., Taylor, S., Crawley, A., Wood, M., 1997. Functional MRI of pain and attention
related activations in the human cingulate cortex. J. Neurophysiol. 77, 3370–3380.
Delaney, A.J., Esmaeili, A., Sedlak, P.L., Lynch, J.W., Sah, P., 2009. Differential expression
of glycine receptor subunits in the rat basolateral and central amygdala. Neurosci.
Lett. 469, 237 –242.
Dhond, R.P., Yeh, C., Park, K., Kettner, N., Napadow, V., 2008. Acupuncture modulates
resting state connectivity in default and sensorimotor brain networks. Pain 136,
Drevets, W.C., 2007. Orbito-frontal structure and function in depression. Ann. NY Acad.
Sci. 1121, 499–527.
Edzard, E.E., Rand, J.I., Stevison, C., 1998. Complementary therapies for depression. Arch.
Gen. Psychiatry 55, 1026–1032.
Fang, S.H., Zhang, S.C., Liu, H.B., 2006. Study on brain responses to acupuncture by
functional nuclear magnetic resonance — observations on 10 healthy subjects.
Zhongguo Zhong xi yi ze he za zhi 26, 965–968.
Fang, J., Zhen, J., Li, K., Kong, J., Nixon, E., Zeng, Y., Tong, H., Wang, P., Hui, K.K.S., 2008.
The salient characteristics of the central effects of acupuncture needling: limbic–
paralimbic–neocortical network modulation. Hum. Brain Mapp. 30, 1196–1206.
Fox, M.D., Snyder, A.Z., Vincent, J.L., Corbetta, M., Van Essen, D.C., Raichle, M.E., 2005.
The human brain is intrinsically organized into dynamic, anti-correlated functional
networks. Proc. Natl. Acad. Sci. USA. 102, 9673–9678.
Fransson, P., 2005. Spontaneous low-frequency BOLD signal ﬂuctuations: an fMRI
investigation of the resting-state default mode of brain function hypothesis. Hum.
Brain Mapp. 26, 15–29.
Fujiwara, S., Sasaki, M., Kanbara, Y., Inoue, T., Hirooka, R., Ogawa, A., 2008. Feasibility of
1.6-mm isotropic voxel diffusion tensor tractography in depicting limbic ﬁbers.
Neuroradiology 50, 131–136.
Golland, Y., Golland, P., Bentin, S., Malach, R., 2008. Data-driven clustering reveals a
fundamental subdivision of the human cortex into two global systems. Neurop-
sychologia 46, 540–553.
Greicius, M.D., Krasnow, B., Reiss, A.L., Menon, V., 2003. Functional connectivity in the
resting brain: a network analysis of the default mode hypothesis. Proc. Natl. Acad.
Sci. USA. 100, 253–
Greicius, M.D., Menon, V., 2004. Default-mode activity during a passive sensory task:
uncoupled from deactivation but impacting activation. J. Cogn. Neurosci. 16,
Greicius, M.D., Flores, B.H., Menon, V., Glover, G.H., Solvason, H.B., Kenna, H., Reiss, A.L.,
Schatzberg, A.F., 2007. Resting-state functional connectivity in major depression:
abnormally increased contributions from subgenual cingulate cortex and thalamus.
Biol. Psychiatry 62, 429–437.
Gusnard, D.A., Raichle, M.E., 2001. Searching for a baseline: functional imaging and the
resting human brain. Nat. Rev. Neurosci. 2, 685–694.
Hagmann, P., Kurant, M., Gigandet, X., Thiran, P., Wedeen, V.J., Meuli, R., Thiran, J.P.,
2007. Mapping human whole-brain structural networks with diffusion MRI. PLoS
ONE 2, e597.
Haker, E.H.E., Bjerring, P., 2000. Effect of sensory stimulation (acupuncture) on
sympathetic and parasympathetic activities in healthy subjects. J. Auton. Nerv. Syst.
Han, J.S., 2004. Min i-revi ew. Acupuncture and endorph ins. Neurosci. Lett. 361,
He, Y., Wang, L., Zang, Y., Tian, L., Zhang, X., Li, K., Jiang, T., 2007. Regional coherence
changes in the early stages of Alzheimer's disease: a combined structural and
resting-state functional MRI study. Neuroimage 35, 488–500.
Holliﬁeld, M., Sinclair-Lian, N., Warner, T.D., Hammerschlag, R., 2007. Acupuncture for
post-traumatic stress disorder: a randomized controlled pilot trial. J. Nerv. Ment.
Dis. 195, 504–513.
Honey, C.J., Sporns, O., Cammoun, L., Gigandet, X., Thiran, J.P., Meuli, R., Hagmann, P.,
2009. Predicting human resting-state functional connectivity from structural
connectivity. Proc. Natl. Acad. Sci. USA. 106, 2035–2040.
Hui, K.K., Liu, J., Makris, N., Gollub, R.L., Chen, A.J., Moore, C.I., Kennedy, D.N., Rosen, B.R.,
Kwong, K.K., 2000. Acupuncture modulates the limbic system and subcortical gray
structures of the human brain: evidence from fMRI studies in normal subjects.
Hum. Brain Mapp. 9, 13–25.
Hui, K.K., Liu, J., Marina, O., Napadow, V., Haselgrove, C., Kwong, K.K., Kennedy, D.N.,
Makris, N., 2005. The integrated response of the human cerebro-cerebellar and
limbic systems to acupuncture stimulation at ST 36 as evidenced by fMRI.
Neuroimage 27, 479–496.
Hui, K.K., Nixon, E.E., Vangel, M.G., Liu, J., Marina, O., Napadow, V., Hodge, S.M., Rosen, B.R.,
Makris, N., Kennedy, D.N., 2007. Characterization of the “deqi” response in
acupuncture. BMC Complement. Altern. Med. 7, 33.
Hui, K.K.S., Marina, O., Claunch, J.D., Nixon, E.E., Fang, J.L., Li, M., Napadow, V., Vangel, M.,
Makris, N., Chan, S.T., Kwong, K., Rosen, B.R., 2009. Acupuncture mobilizes the
brain's default mode and the anti-correlated network in healthy subjects. Brain Res.
Jones, A.K., Brown, W.D., Friston, K.J., Qi, L.Y., Frackowiak, R.S., 1991. Cortical and
subcortical localization of response to pain in man using positron emission
tomography. Proc. Biol. Sci. 244, 39–44.
Kahn, I., Andrews-Hanna, J.R., Vincent, J.L., Snyder, A.Z., Buckner, R.L., 2008. Distinct
cortical anatomy linked to subregions of the medial temporal lobe revealed by
intrinsic functional connectivity. J. Neurophysiol. 100, 129–139.
Kennedy, D.P., Courchesne, E., 2008. The intrinsic functional organization of the brain is
altered in autism. Neuroimage 39, 1877–1885.
Kennedy, D.P., Redcay, E., Courchesne, E., 2006. Failing to deactivate: resting functional
abnormalities in autism. Proc. Natl. Acad. Sci. USA 103, 8275–8280.
Leo, R.J., Ligot Jr., J.S., 2007. A systematic review of randomized controlled trials of
acupuncture in the treatment of depression. J. Affect. Disord. 97, 3–22.
Liang, M., Zhou, Y., Jiang, T., Liu, Z., Tian, L., Liu, H., Hao, Y., 2006. Widespread functional
disconnectivity in schizophrenia with resting-state functional magnetic resonance
imaging. Neuroreport 17, 209–213.
Lin, W.Z., Wang, P., 1994. Experimental Acupuncture. Shanghai Science and Technology
Press, Shanghai, pp. 43–44.
Liu, S., Zhou, W., Ruan, X., Li, R., Lee, T., Weng, X., Hu, J., Yang, G., 2007. Activation of the
hypothalamus characterizes th e response to acupuncture stimulation in heroin
addicts. Neurosci. Lett. 421, 203
Liu, T.T., Shi, J., Epstein, D.H., Bao, Y.P., Lu, L., 2009. A meta-analysis of acupuncture
combined with opioid receptor agonists for treatment of opiate-withdrawal
symptoms. Cell Mol. Neurobiol. 29, 449–454.
Lozano, A., Mayberg, H., Glacobbe, P., Hamani, C., Craddock, R., Kennedy, S., 2008.
Subcallosal cingulate gyrus deep brain stimulation for treatment-resistant
depression. Biol. Psychiatry 64, 461–467.
Lowe, M.J., Mock, B.J., Sorenson, J.A., 1998. Functional connectivity in single and multislice
echoplanar imaging using resting-state ﬂuctuations. Neuroimage 7, 119–132.
Lowe, M.J., Phillips, M.D., Lurito, J.T., Mattson, D., Dzemidzic, M., Mathews, V.P., 2002.
Multiple sclerosis: low-frequency temporal blood oxygen level-dependent ﬂuctua-
tions indicate reduced functional connectivity initial results. Radiology 224,
Lund, I., Lundeberg, T., 2006. Are minimal, superﬁcial or sham acupuncture procedures
acceptable as inert placebo controls? Acupunct. Med. 24, 13–15.
Lundeberg, T., Bondesson, L., Lundström, V., 1985. Relief of primary dysmenorrhea by
transcutaneous electrical nerve stimulation. Acta Obstet. Gynecol. Scand. 64,
Lustig, C., Snyder, A.Z., Bhakta, M., O'Brien, K.C., McAvoy, M., Raichle, M.E., Morris, J.C.,
Buckner, R.L., 2003. Functional deactivations: change with age and dementia of the
Alzheimer type. Proc. Natl. Acad. Sci. USA 100, 14504–14509.
Margolin, A., Avants, S.K., Chang, P., Kosten, T.R., 1993. Acupuncture for the treatment of
cocaine dependence in methadone-maintained patients. Am. J. Addict. 2, 194–201.
Mayberg, H.S., Lozano, A.M., Voon, V., McNeely, H.E., Seminowicz, D., Hamani, C.,
Schwalb, J.M., Kennedy, S.H., 2005. Deep brain stimulation for treatment–resistant
depression. Neuron 45, 651–660.
Napadow, V., Liu, J., Kaptchuk, T., 2004. A systematic study of acupuncture practice:
acupoint usage in an outpatient setting in Beijing, China. Complement. Ther. Med.
Napadow, V., Dhond, R.P., Purdon, P.L., Kettner, N., Makris, N., Kwong, K.K., Hui, K.K.,
2005a. Correlating acupuncture FMRI in the human brainstem with heart rate
variability. Conf. Proc. IEEE Eng. Med. Biol. Soc. 5, 4496–4499.
Napadow, V., Makris, N., Liu, J., Kettner, N.W., Kwong, K.K., Hui, K.K., 2005b. Effects of
electro-acupuncture versus manual acupuncture on the human brain as measured
by fMRI. Hum. Brain Mapp. 24, 193–205.
Napadow, V., Kettner, N., Liu, J., Li, M., Kwong, K.K., Vangel, M., Makris, N., Audette, J.,
Hui, K.K., 2007. Hypothalamus and amygdala response to acupuncture stimuli in
Carpal Tunnel Syndrome. Pain 130, 254–266.
Olausson, H., Wessberg, J., Morrison, I., McGlone, F., Vallbo, A., 2010. The neurophys-
iology of unmyelinated tactile afferents. Neurosci. Biobehav. Rev. 34, 185–191.
Olson, I.R., Plotzker, A., Ezzyat, Y., 2007. The Enigmatic temporal pole: a review of
ﬁndings on social and emotional processing. Brain 130, 1718–1731.
Parvizi, J., Van Hoesen, G.W., Buckwalter, J., Damasio, A., 2006. Neural connections of the
posteromedial cortex in the macaque. Proc. Natl. Acad. Sci. USA 103, 1563–1568.
Rombouts, S.A., Barkhof, F., Goekoop, R., Stam, C.J., Sheltens, P., 2005. Altered resting
state networks in mild cognitive impairment and mild Alzheimer's disease: an fMRI
study. Hum. Brain Mapp. 26, 231–239.
Roschke, J., Wolf, C.H., Muller, M.J., Wagner, P., Mann, K., Grozinger, M., Bech, S., 2000.
The beneﬁt from whole body acupuncture in major depression. J. Affect. Disord. 57,
Schmahmann, J.D., Pandya, D.N., 2006. Fiber Pathways of the Brain. Oxford University
Seeley, W.W., Menon, V., Schatzberg, A.F., Keller, J., Glover, G.H., Kenna, H., Reiss, A.L.,
Greicius, M.D., 2007. Dissociable intrinsic connectivity networks for salience
processing and executive control. J. Neurosc. 27, 2349–2356.
Shi, M.T., Li, L., He, G.R., Lu, G.W., 1986. Effect of somatosensory stimulation at speciﬁc
points on levels of monoamines in cerebrospinal ﬂuid. Acta Beijing Second Med.
Coll. 7, 15–18.
89K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90
Shulman, G.L., Corbetta, M., Buckner, R.L., Raichle, M.E., Fiez, J.A., Miezin, F.M., Petersen,
S.E., 1997. Top-down modulation of early sensory cortex. Cereb. Cortex 7, 193–206.
Snell, R.S., 2006. Clinical Neuroanatomy, 6th edition. Lipincott Williams & Wilkins,
Philadelphia, pp. 142–144.
Stein, E.A., Pankiewicz, J., Harsch,H.H., Cho, J.K., Fuller, S.A., Hoffman, R.G.,Hawkins, M., Rao,
S.M., Bandettini, P.A., Bloom, A.S., 1998. Nicotine induced limbic-cortical activation in
the human brain: a functional MRI study. Am. J. Psychiatry 155, 1009–1015.
Stoffers, D., Bosboom, J.L., Deijen, J.B., Wolters, E.C., Berendse, H.W., Stam, C.J., 2007.
Slowing of oscillatory brain activity is a stable characteristic of Parkinson's disease
without dementia. Brain 130, 1847–1860.
Talairach, J., Tournoux, P., 1988. Co-planar stereotaxic atlas of the human brain: 3-
Dimensional proportional system: an approach to medical cerebral imaging.
Thieme Medical, New York.
Talbot, J.D., Marrett, S., Evans, A.C., Meyer, E., Bushnell, M.C., Duncan, G.H., 1991. Multiple
representations of pain in human cerebral cortex. Science 251, 1355–1358.
Vogt, B.A., 2005. Pain and emotion interactions in subregions of the cingulate gyrus. Nat.
Rev. Neurosci. 6, 533–544.
Vogt, B.A., Berger, G.R., Derbyshire, S.W., 2003. Structural and functional dichotomy of
human mid-cingulate cortex. Eur. J. Neurosci. 18, 3134–3144.
Vogt, B.A., Vogt, L., Laureys, S., 2006. Cytology and functionally correlated circuits of
human posterior cingulate areas. Neuroimage 29, 452–466.
Wang, L., Zang, Y., He, Y., Liang, M., Zhang, X., Tian, L., Wu, T., Jiang, T., Li, K., 2006.
Changes in hippocampal connectivity in the early stages of Alzheimer's disease:
evidence from resting state fMRI. Neuroimage 31, 496–504.
Wang, W., Liu, L., Zhi, X., Huang, J.B., Liu, D.X., Wang, H., Kong, X.Q., Xu, H.B., 2007. Study
on the regulatory effect of electro-acupuncture on hegu point (LI4) in cerebral
response with functional magnetic resonance imaging. Chin. J. Integ. Med. 13,
Wang, H., Qi, H., Wang, B.S., Cui, Y.Y., Zhu, L., Rong, Z.X., Chen, H.Z., 2008. Is acupuncture
beneﬁcial in depression: a meta-analysis of 8 randomized controlled trials. J. Affect.
Disord. 111, 125–134.
Wedeen, V.J., Wang, R.P., Schmahmann, J.D., Benner, T., Tseng, W.Y., Dai, G., Pandya, D.N.,
Hagmann, P., D'Arceuil, H., de Crespigny, A.J., 2008. Diffusion spectrum magnetic
resonance imaging (DSI) tractography of crossing ﬁbers. Neuroimage 41, 1267–1277.
Willis, W.D., 1989. The origin and de stination of pathways invo lved in pain
transmission, In: Melzack, R., Wall, P.D. (Eds.), Textbook of Pain, 2nd ed. Churchill
Livingstone, Edinburgh, pp. 112–127.
Willis, W.D., Westlund, K.N., 1997. Neuroanatomy of the pain system and of the
pathways that modulate pain. J. Clin. Neurophysiol. 14, 2–31.
Wu, M.T., Hsieh, J.C., Xiong, J., Yang, C.F., Pan, H.B., Chen, Y.C., Tsai, G., Rosen, B.R.,
Kwong, K.K., 1999. Central nervous pathway for acupuncture sti mulation:
localization of processing with functional MR imaging of the brain-preliminary
experience. Radiology 212, 133–141.
Yang, C.H., Shim, I., Kim, S.Y., 2001. Acupuncture attenuates the neurochemical and
behavioral sensitization to repeated DA in cocaine treatment. Abstr. Soc. Neurosci.
Yoon, S.S., Kwon, Y.K., Kim, M.R., Shim, I., Kim, K.J., Lee, M.H., Lee, Y.S., Golden, G.T., Yang,
C.H., 2004. Acupuncture-mediated inhibition of ethanol-induced dopamine release
in the rat nucleus accumbens through the GABA
receptor. Neurosci. Lett. 369,
Zang, Y.F., He, Y., Zhu, C.Z., Cao, Q.J., Sui, M.Q., Liang, M., Tian, L.X., Jiang, T.Z., Wang, Y.F.,
2007. Altered baseline brain activity in children with ADHD revealed by resting-
state functional MRI. Brain Dev. 29, 83–91.
Zhou, Y., Liang, M., Jiang, T., Tian, L., Liu, Y., Liu, Z., Liu, H., Kuang, F., 2007. Functional
dysconnectivity of the dorsolateral prefrontal cortex in ﬁrst-episode schizophrenia
using resting-state fMRI. Neurosci. Lett. 417, 297–302.
90 K.K.S. Hui et al. / Autonomic Neuroscience: Basic and Clinical 157 (2010) 81–90