GASTROENTEROLOGY 2006;131:19251942

REVIEWS IN BASIC AND CLINICAL GASTROENTEROLOGY

Neuroimaging of the Brain-Gut Axis: From Basic Understanding to Treatment of Functional GI Disorders
EMERAN A. MAYER,*, BRUCE D. NALIBOFF,*,, and A. D. BUD CRAIG
*Center for Neurovisceral Sciences & Womens Health; CURE: Digestive Disease Research Center; David Geffen School of Medicine at UCLA, Los Angeles, California; VA Greater Los Angeles Healthcare System, Los Angeles, California; and Barrow Neurological Institute, Phoenix, Arizona

We are enthusiastic in the introduction of a new monthly review article series, entitled Reviews in Basic and Clinical Gastroenterology. Written by authorities in their respective elds, the objective of each review article is to provide an overview of a particular theme or topic for the broad scientic and clinical readership. Within a given topic, both basic and clinical aspects will be covered, accompanied by key gures and relevant references. The reader will also appreciate that topics will be interwoven from month to month as well. We hope you enjoy this section.David C. Metz, MD, Wak El-Deiry, MD, PhD, and Anil K. Rustgi, MD

ltered reex and perceptual responses within the brain-gut axis have emerged as a generally accepted model to explain the cardinal symptoms of functional gastrointestinal (GI) syndromes. The ability to image the living human brain with various neuroimaging modalities has greatly enhanced our ability to study these brain gut interactions in health and disease. Reex responses within the brain-gut axis, mediated by lamina I and vagal afferents, and efferents of the autonomic nervous system, play a crucial role in the maintenance of homeostasis during physiological perturbations caused by food intake, contractile activity, and metabolic products of the enteric ora. The insular cortex plays an important role in the conscious perception of all sensations arising from the body, while the dorsal anterior cingulate cortex (dACC), with its connection to effector systems, mediates the affective response and motivational drive. The magnitude and gain of these processes is highly inuenced by central arousal systems and top-down corticolimbic modulation, mediating the effect of environmental context, emotions, cognitions, and memories on perception and gut function. The majority of neuroimaging studies of the brain gut axis in humans are consistent with the model of parallel processing of afferent information in insula and dACC. Newer hypothesis driven studies, studying the differential contributions of afferent input, central arousal systems, and cortico-limbic pontine interactions have greatly contributed to our understanding of brain gut interactions in health and disease. fMRImediated detection of changes in the activity and connectivity of brain regions involved in these different processes by pharmacological and behavioral therapies holds great promise for the development of novel approaches to functional GI disorders.

viscera, skin, muscle, joint, and teeth. All of these bers signal changes in the physiological condition of the body and provide the essential sensory input that is crucial for the autonomic responses that maintain homeostasis.1

Homeostatic Afferent Processing Network

The term refers to a brain network that is consistently activated in response to homeostatic afferent ber activation. This network includes the insular and dACC, thalamic nuclei (MDvc, VMb, VMpo), and the parabrachial nucleus (PBN). Since non-painful and painful visceral and somatic stimuli, as well as emotional stimuli, can activate this network, the term pain matrix commonly used to describe these regions in the literature no longer seems appropriate.

Feelings
Several specic meanings are used for this word in common usage. First, we experience different feelings from our bodiesincluding satiety, abdominal pain, and discomfort which represent afferent sensory input from receptors and can be regarded as sensations (eg, visceral sensations). Second, we experience feelings associated with our ongoing emotional condition, otherwise known as mood or affective state, such as anxiety, contentment, or irascibility. We include also what we
Abbreviations used in this paper: dACC, dorsal anterior cingulate cortex; DNIC, diffuse noxious inhibitory control; FD, functional dyspepsia; fMRI, functional magnetic resonance imaging; IGLEs, intraganglionic laminar endings; LCC, locus coeruleus complex; MDvc, ventral caudal portion of the medial dorsal nucleus (of the thalamus); NCF, nucleus cuneiformis; NTS, nucleus tractus solitarius; pACC, perigenual cingulate cortex; PAG, periaqueductal gray; PBN, parabrachial nucleus; PET, positron emission tomography; PFC, prefrontal cortex; VMb, basal part of the ventral medial nucleus; VMpo, posterior part of the ventral medial nucleus (of the thalamus). 2006 by the AGA Institute 0016-5085/06/$32.00 doi:10.1053/j.gastro.2006.10.026

Glossary Homeostatic Afferents

The term refers to small-diameter sensory afferent bers terminating in lamina I of the spinal cord that innervate all of the tissues and organs of the body, including the

1926

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

regard as homeostatic feelings, such as chilliness, achiness, or burning pain, that represent our physical condition. Third, we experience feelings associated with strong emotions elicited by social or environmental conditions, such as anger, sadness, happiness, and so on, and the evaluation of such conditions. These feelings in particular represent the awareness of our behavioral condition. The subjective experience of all of these types of feelings is completely dependent on self-awareness in humans.2,3

Emotion
The term is used to describe a neurobehavioral state adapted for the attainment of a particular goal or the resolution of particular conditions as described by Rolls.4 It is characterized in humans as a motivation accompanied by a characteristic feeling and autonomic sequelae. Emotional behavior may occur without awareness or without a concomitant feeling, as during unconscious emotional actions, or as in animals that do not display self-awareness. Emotions are viewed as ongoing and continuously varying events.

Homeostatic Emotions
We dene these as the motivations and feelings that are associated with changes in the bodys physiological condition and with the autonomic responses and behaviors that occur in order to restore an optimal balance.2 For example, if your body is hypoglycemic, you feel hunger and you are motivated to eat. Homeostatic emotions are the background emotions that affect our energy level, our mood, and our disposition. Spinal and vagal visceral afferent input to the central nervous system plays an important role in the generation of such emotions.

in brain imaging studies of visceral pain (in particular insula, dACC, and prefrontal cortex [PFC] regions), regardless of experimental paradigm, as well as in studies using somatic or emotional stimuli. Finally, there is a need for more hypothesis driven experimental designs to deconstruct brain circuits activated by visceral afferent stimuli into afferent processing networks, as well as overlapping networks activated by cognitive, emotional, and arousal components of the experimental setting. Such identication of specic brain circuits is the prerequisite for evaluation of pharmacologic and behavioral therapies using imaging technologies (for details, see part II). In the current review, we will discuss brain-gut interactions in health and disease based on the concept of homeostatic reex regulation, the conscious perception of homeostatic responses, and the possible alteration of both processes in functional GI disorders. We will provide an overview of the functional neuroanatomical basis for this concept derived from preclinical research, and we will review published neuroimaging studies performed in humans which are consistent with many of the functional neuroanatomical concepts laid out in the rst part of this review. Finally, we will discuss implications of this conceptualization for the development and evaluation of novel therapeutic approaches to functional GI disorders.

The Concept of Homeostatic Emotions

It has traditionally been taught that the gut and the skin differ starkly in their relationship with the brain.10 This seems intuitively acceptable, because there are major differences: the gut and the skin certainly feel different; we can localize sensations on our skin, but we have difculty pinpointing sensations from our viscera and generally refer them to somatic dermatomes; we have voluntary control over our limbs, but very little control over our gut. Nevertheless, recent neurobiological evidence suggests that there is a common pattern in the neural connections of all tissues of the body gut, skin, muscle, joint, teeth, and so onthat is evolutionarily related to the primal need for the brain to maintain the integrity of the entire body. This process is called homeostasis. The pathways for homeostatic control of the body are evolutionarily ancient. Each of the various tissues of the body generates sensory inputs that signify changes in physiological conditions, and the brain integrates these signals to maintain a homeostatic balance of the entire body that optimizes the use of energy for survival. We humans uniquely experience feelings from each of the tissues of our bodies because evolution has produced brain mechanisms for the conscious perception of an image of the physiological condition of our bodies (ie, how you feel). In the healthy person, and in the absence of arousal and attentional focus on a particular feeling (such focus does occur with painful stimulus intensities), these feelings and sensations contribute to background emotions3 but do not distract from constant cognitive demands on the individual. In contrast, in the functional GI patients, there is an alteration in the attentional, perceptual, and affective response to such feelings from the digestive tract, typied by such expressions as I feel constipated, I feel bloated, I dont feel right. The classical term interoception, formerly used only to refer to visceral sensation, has recently been redened, based on these new ndings, to refer to the sense of the physiological condition of the body. The new neuroanatomical ndings indicate that all of the feelings from our bodies reect its physiological condition and

I. Introduction
Bidirectional brain-gut interactions play an important role in the regulation of many vital functions in health and disease. In health, brain-gut interactions play a crucial role in the regulation of digestive processes (including the regulation of food intake and bowel movements), in the modulation of the gut-associated immune system, and in the coordination of the overall physical and emotional state of the organism with digestive processes (reviewed in Mayer and Saper 20005). Altered brain-gut interactions are likely to underlie symptom generation in several functional GI disorders, including irritable bowel syndrome (IBS), functional dyspepsia (FD), non-cardiac chest pain, and cyclical vomiting syndrome. Although less well characterized, alterations in brain-gut interactions may also be involved in the modulation of disease activity in inammatory bowel disease (reviewed in Mayer 20006) and in the pathophysiology of various eating disorders.7 Although the term brain-gut axis provides a general construct to explain how alterations in perception and in the autonomic regulation of the gut contribute to the cardinal symptoms of abdominal pain and discomfort reported by functional GI patients, the concept does not provide a comprehensive framework to understand other characteristic features of these disorders, such as the overlap between different functional GI syndromes,8 and the overlap between functional GI syndromes with other visceral, somatic, and affective disorders (reviewed in Wesseley et al 19999). A broader framework is also needed to interpret the consistent widespread activations seen

December 2006

THE BRAIN-GUT AXIS

1927

can be viewed as homeostatic emotions2 or background emotions.3 This includes such typical visceral sensations as hunger, thirst, vasomotor ush, satiation, fullness, and urgency, but also a range of other sensations such as temperature, itch, muscle ache, and pain. Like all emotions, these comprise both a feeling (sensation) and a motivation (affect). This is particularly relevant in the context of functional GI disorders, where symptoms are not limited to discomfort and pain attributed to the viscera, but frequently include a whole range of other feelings of physical and emotion discomfort.9 The phylogenetically new pathways in human and primate brains generate both of these aspects directly, resulting in forebrain activation patterns that resemble the activation pattern common to all human emotions (that is, activation of the anterior insula and dACC). The redened concept of interoception emerges directly from the functional anatomical data summarized below. This concept is strongly reminiscent of the natural philosophical concept of Gemeingefhl (common sensation) in the German literature of the 1800s, of the sense of the material me described by Sherrington in 1900, and of the idea that human emotions are based on feelings from the body, which is the essence of the James-Lange theory of emotion and its recent reformulation by Damasio.3 In order to understand how these ndings relate feelings from the gut to homeostasis and how homeostatic regulation controls feelings from the gut, the nature of homeostasis and homeostatic emotions must rst be described.

motivation (and the accompanying autonomic changes) is generated in non-humanoid mammals by a signal to the forebrain (ie, the behavioral controller) from the main homeostatic afferent integration site in the brainstem, the parabrachial nucleus (PBN). In humans, evolution produced a new direct (spinothalamo-cortical) pathway to the forebrain that surmounts the primal pathway and which generates both an affective motivation and a sensation in the limbic motor and sensory cortices (Figure 1A). The basic homeostatic (interoceptive) feelings, or modalities, not only include hunger, thirst, stomach cramps, fullness, and rectal urgency, but all feelings from the body such as temperature, itch, and muscle ache. So, consistent with the view that an emotion in humans consists of both a sensation and a motivation with direct autonomic sequelae,4,15,16 these feelings are the human percepts of distinct homeostatic emotions that directly relate the bodys needs. As will be discussed below, the concept of homeostatic emotions, regardless of the valence of the emotion has direct relevance to the interpretation of ndings in human neuroimaging studies of visceral stimuli. It is consistent with the reported activation of similar brain regions (insula, dACC) by a variety of both pleasant and unpleasant stimuli (reviewed in Vogt 200517).

Pain and Other Feelings From the Body as Aspects of Homeostasis

Pain, whether from the skin or from the gut, is often regarded as a distinct feeling, but from the perspective laid out above, pain can be viewed as another homeostatic feeling. It has characteristics exactly comparable to other feelings from the body. Pain normally originates with a change in the condition of tissues, a physiological imbalance that automatic (subconscious) homeostatic systems alone cannot rectify. It comprises both a sensation and an affective behavioral drive with accompanying autonomic adjustments. Depending upon conditions, pain can be unpleasant (as usual) or pleasant (such as when it relieves an intense itch). Pain also generates characteristic reexive motor patterns, as do itch, hunger, and experimental gut stimuli. The behavioral motivation of pain is normally correlated with the intensity of the sensory input, but this can vary under different behavioral, autonomic, and emotional conditions, so that pain can become intolerable or it can disappear, similar to any other homeostatic emotion (eg, hunger). The modulation of the motivational aspect of a homeostatic emotion like pain can occur via inhibitory or excitatory prefrontal inuences on brain circuits involving the dACC.18 20 However, unremitting pain that outlasts its homeostatic role is pathological.21 Viewing pain and other feelings from the body as homeostatic emotions provides a ready explanation of the interactions of these feelings from the body, with other homeostatic conditions (including the level of arousal, mood, and affect) since homeostasis is an integrated, dynamic process. This conceptual perspective also provides a rm basis for explaining the interactions of pain with emotional status or attention (ie, the psychological dimension of pain), and it unies the different conditions that can cause different types of pain from different tissues under a common homeostatic functionthe maintenance of the integrity of the body. All animals respond with emotional behavior (including musculoskeletal and autonomic responses) to stimuli that in humans cause a feeling of pain.16 An example relevant to preclinical studies in visceral pain mech-

Homeostatic Emotions Drive Behavior

The interplay between the feeling and the motivation of a homeostatic emotion has been studied in great detail with thermal stimuli,1,2,1114 but can easily be adapted to the stimuli relevant to the digestive system. For example, the affect (pleasantness, unpleasantness) experienced in connection with food stimuli is the perceptual correlate of obligatory motivations. In the prolonged absence of food and water needed to maintain metabolic balance, a person will experience increasing discomfort (even hunger pains). If kept in this situation long enough, the discomfort grows until it becomes an intractable motivation. When the person is nally able to eat and drink, it feels pleasant, and he/she is motivated to eat and drink until no longer hungry. Just as a cool stimulus feels wonderful if you are hot, but unpleasant if you are cold, so too, under opposing conditions of hunger or satiety, the affect generated by the availability of food and water strikingly invertsthus, a large intake of food feels wonderful if you are hungry, but the same meal can produce an unpleasant feeling of nausea if you are already full. In the same way, eating salt or sugar is pleasant (and thus motivated) if the body needs it, but unpleasant if it doesnt (so-called stimulus-specic satiety4). In other words, the specic affective feeling that we perceive reects the metabolic (homeostatic) behavioral motivation that originates from the bodys needs. Sensory inputs that relate the bodys physiological condition, or homeostatic afferents, drive the homeostatic mechanisms that promote survival. The primordial means of regulating hunger, thirst, abdominal fullness, and the evacuation of waste products (stool and urine) in all vertebrates is motivated behavior, so the pathways that guide homeostatic sensory input to motivational processes must be ancient. Such behavioral

1928

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

Figure 1.

Ascending projections of homeostatic afferents. (A) Organization of interoceptive pathways. Small diameter afferents that travel with sympathetic and with parasympathetic efferents provide input to lamina I and NTS, respectively. In mammals, the activity of both types of afferents is integrated in the PBN, which projects to insular cortex. In non-human and human primates, a direct projection from lamina I and from the NTS exist to ventromedial thalamic nuclei (VMpo and VMb, respectively). Neurons in these nuclei project in a topographical fashion to the mid/posterior insula. In humans, this cortical image of the homeostatic state of the organism is re-represented in the anterior insula on the same side of the brain. These re-representations provide the substrate for a subjective evaluation of interoceptive state. PBN, parabrachial nucleus. Reprinted with permission from Craig 2002.2 (B) Spino-thalamo-cortical system. Summary diagram illustrating the projections in primates of homeostatic afferent pathways from lamina I (spinal) and NTS (vagal) to thalamic nuclei, and the 2 cortical regions involved in the sensory (insula) and motivational (ACC) dimensions of homeostatic emotions. NTS, nucleus tractus solitarius. ACC, anterior cingulate cortex. Modied from Craig 2003.138

anisms is the so-called visceromotor reex seen in response to colorectal distention in rodents.22 The new data described in the following text reveal that noxious stimuli, whether from the viscera or from the skin, are represented in an evolutionarily ancient neural pathway that has the primary purpose of driving homeostatic mechanisms at the enteric (for gut stimuli), spinal, and brainstem levels and generating integrated behavioral motivation at the forebrain level. In primates, novel thalamocortical projections have emerged from this basic homeostatic system that provides direct pathways to encephalized cortical mechanisms for highly resolved sensations and motivations. Notably, sub-primates have only the subcortical mechanisms that drive integrated homeostatic (emotional) behavior, that is, they do not have these direct telencephalic pathways that produce an interoceptive cortical representation of the precise

physiological condition of the body in humans. In humans, these novel pathways are further elaborated, re-represented, and integrated with other forebrain emotional components in the anterior insula and orbitofrontal cortex (the fronto-insular region) so that a conscious perception of the entire emotional moment is formed that is based, evolutionarily and functionally, on the bodys homeostatic condition, including the state of the gastrointestinal tract (Figure 1AB). As will be discussed in the clinical part of this review, this concept has considerable implications for the study of mechanisms underlying visceral pain and autonomic dysregulation in patients with functional GI disorders, using neuroimaging approaches. In the majority of these patients (as well as patients with a wide range of other functional disorders), the primary symptoms of abdominal pain and discomfort are directly re-

December 2006

THE BRAIN-GUT AXIS

1929

Figure 2. Hierarchical organization of homeostatic reex systems involving the sympathetic nervous system. (A) Homeostatic afferents that report the physiological condition of all tissues in the body, including the GI tract, terminate in lamina I of the dorsal horn. The ascending projections of these neurons provide the basis for reex arcs at the spinal, medullary and mesencephalic levels. Limbic, paralimbic, and prefrontal centers provide modulatory inuences on the gain of these reexes. Reprinted with permission from Craig 2002.2 (B) Cortical modulation of homeostatic afferent input to the central nervous system. PFC regions (dorsolateral PFC [dlPFC], orbitofrontal cortex [orbFC]) modulate activity in limbic and paralimbic regions (amygdala [amy], ACC subregions, and hypothalamus [Hypoth]), which in turn regulate activity of descending inhibitory and facilitatory descending pathways through the PAG and pontomedullary nuclei. Activity in these corticolimbic pontine networks mediates the effect of cognitions and emotions on the perception of homeostatic feelings, including visceral pain and discomfort.

lated to the altered sensory and affective perception of homeostatic feelings associated with the esophagus, stomach, or intestine (bloating, urgency, pain). In addition, altered homeostatic reex responses may contribute to ndings such alterations in regional gastrointestinal transit, and in reex responses affecting gastrointestinal motility and secretion. Since rodents, the animals most commonly used in experiments to model these disorders, do not have the forebrain structures to generate the conscious emotional feelings of humans, ndings obtained in such animals (using so-called pseudo-affective reex responses) may reect primarily the phylogenetically shared enteric, spinal and brainstem components of homeostatic pathways (eg, reexes), but may provide less insight into the uniquely human experience of abdominal pain and discomfort. This limitation may be particularly important for disease relevant cortical modulation mediating cognitive and emotional inuences on the experience of visceral sensations. It is conceivable that the relatively low yield of conventional drug discovery strategies that have heavily relied on testing in animal models to produce effective medications for functional GI disorders, may be related to this fundamental difference in the

conscious experience of homeostatic feelings between humans and non-humanoid animals.

II. Basic Aspects The Functional Neuroanatomy of Homeostatic Afferent Pathways

Small-diameter primary afferent bers that report the physiological status of all of tissues (including nociceptors, thermoreceptors, osmoreceptors, and metaboreceptors) terminate monosynaptically on projection neurons in lamina I of the spinal dorsal horn. Developmentally, these small-diameter afferents originate from a second wave of small dorsal root ganglion cells that emerge subsequent to the large cells that generate mechanoreceptors and proprioceptors, and their projection into the dorsal horn is temporally coordinated with the appearance of lamina I neurons.23 The lamina I neurons originate from progenitors of interneurons in the lateral horn (the sympathetic cell column) and migrate to the top of the dorsal horn (aided by a ventromedial rotation of the entire dorsal horn) precisely at the right time to meet the incoming small

1930

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

diameter afferents. This coordinated development indicates that together the small-diameter afferents and the lamina I neurons constitute a cohesive system for homeostatic afferent activity that parallels the sympathetic nervous system. The small-diameter afferents in cranial parasympathetic nerves (eg, vagus, glossopharyngeal) terminate similarly in the medullary nucleus of the solitary tract (NTS). Notably, the small-diameter afferents report the physiological status of all tissues of the body, including viscera, muscle, as well as skin (the largest organ of the body), so that these are not simply spinal and vagal visceral afferents,24 but rather can be called homeostatic afferents. The small-diameter afferent bers that innervate the gut course peripherally with sympathetic and parasympathetic nerves (eg, splanchnic, pelvic, vagus), and with somatic nerves (eg, pudendal). Many innervate vessels in the mesentery and have a cardiovascular role, while others innervate the mesenteric ganglia and play a role in the communication between the central nervous system and the enteric nervous system, the virtually autonomous component of the autonomic nervous system that controls the gut. Many respond to chemical and mechanical stimuli over a broad range of thresholds and response slopes, so that they might be considered to represent a broad continuum of response properties. However, they can also be regarded as separable into different classes, including low-threshold, high-threshold (nociceptive), silent (unresponsive unless inamed), and thermosensitive (primarily warm25,26). These classes may correspond to different functional roles and different subjective feelings. Within the gut, there are also several distinct anatomical patterns of innervation, with some bers ending within the longitudinal muscle sheets, others ending within the enteric ganglia (intraganglionic laminar endings, IGLEs),27 and others in close proximity to enterochromafn28 or to mast cells.29 Recent evidence from in vitro experiments suggests that these different terminal morphologies correspond to different patterns of physiological sensitivity.30,31 The likelihood that distinctly different physiological classes of visceral afferents exist and activate different central pathways is supported by anatomical evidence showing that different neurons in the supercial dorsal horn are activated by renal artery occlusion and by renal vein occlusion, probably by responding selectively to renal osmoreceptors or mechanoreceptors, respectively.32 The silent ber type may be of particular interest with respect to clinical disorders, because such bers from the skin (also called sleeping or mechanically insensitive) have recently been identied as a distinct biological category that is directly responsible for the central sensitization caused by peripheral inammation.33 Notably, most of these ndings stem from observations of afferent bers that parallel the sympathetic system; vagal afferent bers are less well studied, but include bers innervating the gut that are metaboreceptive and chemoreceptive and are likely associated with the ingestive and nutritive functions of the gut.27,34,35

meostasis.1,2 In the spinal cord, their only major projection is to the sympathetic cell column of the thoracolumbar spinal cord, where autonomic pre-ganglionic output neurons are located. In the brainstem, they project exclusively to the recognized homeostatic integration sites (eg, caudal and rostral ventrolateral medulla, catecholamine cell groups A1-2 and A5-7, PBN, periaqueductal gray [PAG]), which also receive parasympathetic afferent activity by way of the NTS and which are heavily interconnected with the hypothalamus and amygdala. The NTS that receives small-diameter bers from parasympathetic nerves similarly projects to all of these sites. These spinal and bulbar projections from lamina I and from the NTS provide the substrate for the hierarchical, modality-selective somato-autonomic reexes activated by spinal small-diameter afferents that are crucial for homeostatic control of all tissues of the body.36 The lamina I spino-bulbar projection includes all known lamina I cell types, reecting input from all physiological classes of homeostatic afferent activity. Notably, these are not just emergency pathways, but are engaged in an ongoing basis; for example, respiration is linearly modulated by skin temperature,37 and cardiorespiratory activity is directly affected by ongoing muscular activity.11,38

Spino-Thalamo-Cortical Pathway in Humans

Ascending lamina I afferent activity is integrated mainly in several brainstem sites (A1, PBN, PAG) in sub-primates, which then provide an integrated signal to behavioral control regions in the forebrain. In primates, however, there is a novel, additional lamina I STT projection to a specic thalamo-cortical relay nucleus (VMpo) in posterolateral thalamus, which in turn projects to a discrete portion of dorsal posterior insular cortex (Figure 1).2 This interoceptive cortex contains modalityselective representations of all afferent activity from lamina I (ie, sympathetic afferent input) and from the NTS (ie, parasympathetic input). In monkeys, this pathway is just visible, but in humans, it is greatly enlarged. Many functional imaging studies in humans conrm its role in pain, hunger, thirst, temperature, itch, muscle sensation, sensual touch, and cardiorespiratory activity, etc. (reviewed in Craig 20022), and it can be regarded as primary sensory cortex for the physiological condition of the body in primates. By contrast, in sub-primates the insular cortex appears to have a primordial role in modulating brainstem homeostatic integration (in PBN and other sites) and shows convergent, nonselective responses to homeostatic afferent inputs. The ascending afferent lamina I pathway in primates and humans also provides a direct thalamo-cortical pathway (by way of MDvc in medial thalamus) that activates the dACC (Figure 1). In sub-primates, the dACC receives only integrated homeostatic input from the brainstem.2,39 On the basis of functional imaging and lesion studies in humans, the dACC can be directly associated with the affective aspect of visceral and somatic pain (unpleasantness as well as pleasantness), and with volition, behavioral motivation, and autonomic responses.2,19,39 41 Its interconnections with PFC regions, insular, and ventral striatal regions, along with its strong descending projections to the brainstem, particularly the PAG, strongly support the idea that it can be regarded generally as the limbic behavioral motor cortex, just as the insula can be regarded as the limbic sensory cortex.17,42

Hierarchical Organization of Homeostatic Reexes

The lamina I neurons that receive the small-diameter ber inputs have projections within the spinal cord and brainstem (Figure 2A) that clearly reveal their direct role in ho-

December 2006

THE BRAIN-GUT AXIS

1931

ing that natural homeostatic means for modulation of stressrelated exacerbation of painful syndromes, such as relaxed, slow breathing, may be based on the neuroanatomy of homeostatic emotions.43

Descending Modulation of Homeostatic Reexes and Feelings

Lamina I neurons not only receive input via somatic and visceral afferents, but also receive descending (facilitatory and inhibitory) modulation directly from brainstem pre-autonomic sources, including serotonergic nuclei within the rostral ventral medulla and the noradrenergic pontine locus coeruleus complex (LCC; Figure 2); indeed, lamina I and the spinal autonomic columns are the only regions in the spinal cord that receive descending modulation from the hypothalamus. Thus, the activity of lamina I neurons that ultimately produces the various feelings from the body is modulated by various tonic and phasically active, descending inhibitory, and facilitatory pathways44 46 whose primary purpose is control of homeostatic integration. Recent evidence (in rats) suggests that the gain of the spino-bulbo-spinal reex loops originating from neurokinin 1-receptor-containing lamina I neurons is not constant but can be modulated by peripheral primary afferent inputs from inamed tissue or damaged nerves,47 adapting the homeostatic response to the overall state of the organism. Thus, since the descending forebrain control of these pre-autonomic brainstem regions originates in brain networks associated with attention, emotion generation (ie, the limbic system), and emotion regulation (PFC), these anatomic connections provide the basis for corticolimbic modulation of the afferent activity associated with feelings from the body, including pain, at the spinal level as well as at the brainstem and forebrain levels (Figure 2).

Figure 3. Schematic illustration of different, overlapping brain networks mediating the effects of cognitions and emotions on the perception of homeostatic feelings, including visceral pain and discomfort. Differential dysregulations of one or several of these networks could result in altered perception, even in the presence of normal visceral afferent input to the brain.

The VMpo also has a collateral projection to a portion of sensorimotor cortex (area 3a) that is intercalated between the primary somatosensory area and the primary motor area. Similarly, vagal afferent activation occurs in the lateral portion of area 3a of the primary sensorimotor region by way of the NTS and VMb. These projections can be associated with cortical control of the reex actions of skeletal muscle in response to homeostatic afferent inputs, a role subsumed under the term viscero-somatic integration. The forebrain interoceptive representation of the bodys condition in the dorsal posterior insula is successively rerepresented in the middle insula and then in the right (nondominant) anterior insula.2 Functional imaging data show that the right anterior insula is associated with subjective awareness of homeostatic emotions (eg, visceral and somatic pain, temperature, sexual arousal, hunger, and thirst) as well as all emotions (eg, anger, fear, disgust, sadness, happiness, trust, love, empathy, social exclusion). This region is intimately interconnected with the dACC, which is co-active in such studies. Thus, the meta-representation of the state of the body in the right anterior insula can be regarded as an image of the material self as a feeling (sentient) entity that engenders emotional awareness, consistent with the JamesLange theory of emotion and with the somatic marker hypothesis of Damasio.3 Thus, viewing feelings from the body as homeostatic emotions enables neuroanatomical explanations of interactions with other homeostatic functions and with emotions at the level of the forebrain. Finally, recent evidence showing asymmetry in the sympathetic and parasympathetic afferent re-representations and control mechanisms in the left and right insula and ACC seems to accord with the psychophysiological evidence for the forebrain asymmetry of positive and negative emotions, suggest-

III. Clinical Aspects

Visceral pain and discomfort, including hunger, fullness, early satiety, nausea, bloating, or abdominal pain in humans is a subjective, conscious experience, that results from the modulation of homeostatic feelings by cognitive (attention, expectation), emotional (arousal, anxiety), and motivational factors, as well as memories of past experiences. Thus, the conscious experience is an image of the homeostatic state of the body represented in the insular cortex (Figure 1) and is further modied by these cortical and limbic inputs. In principal, altered perception of visceral stimuli could result from activity changes in visceral afferent signal processing areas alone (reecting increased visceral afferent input to the brain from the gut), from alterations in distinct but overlapping pain modulation circuits (central pain amplication), or from variable combinations of these 2 overlapping circuitries48 50 (Figure 3).

Activation of Regions Involved in Homeostatic Emotions in Studies of the Human Brain-Gut Axis
Following the initial publication of a manuscript describing brain responses to expected and delivered rectal distention in humans in 1997,51 there has been a series of studies describing brain regions activated during brief, experimental, visceral stimulation. The majority of these studies to date have used distention of the rectosigmoid colon, esophagus, and stomach. In addition, a few studies have been reported using

1932

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

chemical-induced discomfort in the esophagus.52,53 Derbyshire, reviewing 15 relevant articles with 21 independent study samples51,52,54 66 published a comprehensive review of these studies up to May of 2002.67 The studies included positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) assessments of an active stimulation condition compared with a non-distention or reduced stimulation baseline condition. Overall, the regions reported as being activated by the experimental visceral stimuli were comparable with those reported in studies of noxious somatic stimulation.68 70 Consistent with the functional neuroanatomy of lamina I afferents outlined earlier, the review identied the insular cortex as the single most consistently activated brain region. In addition, a majority of studies also reported activation of cortical regions including the dACC, the primary sensory cortex, and PFC regions. Although activation of these areas were found for both upper and lower GI stimulation, lower GI stimulation was more consistently associated with activation of more rostral portions of the dACC that have direct connections with limbic and brainstem structures (including the PAG and amygdala) involved in autonomic regulation, while upper GI stimulation (primarily of the esophagus) was more consistently associated with activation of areas involved in sensory and motor processing including more posterior aspects of the dACC (eg, the midcingulate cortex). Two studies have looked at sex differences in response to rectal stimuli in healthy subjects. Kern et al studied brain responses in 28 healthy control subjects (15 women) to aversive rectal distention using fMRI.66 Subjects received individualized distention intensities that were either below (subliminal) or just above the perception threshold. In both sexes, increasing stimulus intensity was associated with increases in brain activation. Volume of cortical activity during distention was signicantly greater in women at all distention levels. Men showed localized clusters of fMRI activity primarily in the sensory motor cortex and parieto-occipital regions, whereas women also showed activity in the insular cortex, dACC, and PFC regions. Berman et al71 studied brain responses in 13 healthy adults (6 women) during 15 s of cued rectal distention at 2 pressures: 25 mm Hg (uncomfortable), and 45 mm Hg (mild pain), as well as during an expectation condition (no distention). The 45-mm Hg pressure signicantly activated the insula and dACC in both sexes. When the number of activated voxels, number of deactivated voxels, and ratio of deactivated voxels to total voxels affected were assessed via random effects mixed-model analyses combining subject data at the region level, greater insula activation in men compared with women was seen during the expectation condition and during the 25 mm Hg distention. In contrast, greater deactivations in women were seen in the amygdala (25 mm Hg distention) and midcingulate (45 mm Hg distention). Women had a signicantly higher proportion of deactivated voxels than men in all 4 subcortical structures during the 25 mm Hg distention. In summary, studies published during the past 5 years using widely different experimental paradigms have conrmed the consistent activation of the central homeostatic afferent processing network consistent with the robust activation of visceral afferent pathways despite varying experimental paradigms and analysis techniques.53,7276 By contrast, the variable activations of PFC and limbic regions, and the seemingly contradictory results on sex differences in reported studies is reective of the

fact that experimental paradigms and analysis techniques varied, and that the majority of these studies did not take into account cognitive and emotional aspects of pain modulation.

Similarities and Differences in Brain Responses to Visceral and Somatic Pain Stimuli
Consistent with the concept of homeostatic emotions generated by visceral and somatic afferent stimulation, similar activation patterns of anterior insula and dACC in response to somatic (generally cutaneous heat or pressure) and visceral stimuli (balloon distention, acid perfusion) have been reported, even though some differences in the relative amounts of activation and location of activations within these regions was observed.67 Several studies have directly compared the brains response to visceral and somatic pain stimuli. For example, Bushnells laboratory has examined the perceptual and central nervous system responses to visceral and cutaneous painful stimuli matched in terms of perceived intensity within the same chest dermatome.7779 The authors found that the visceral, mechanical stimulus was experienced as more unpleasant, diffuse, and variable than cutaneous thermal pain of similar intensity, independent of the duration of the stimulus.77 The cutaneous heat pain evoked higher activations in the bilateral anterior insular cortex and ventrolateral PFC. However, pain in the same dermatome from distention of the esophagus was associated with activation of bilateral inferior primary somatosensory cortex, bilateral primary motor cortex and a more rostral region within the dACC. Evaluating differential brain responses to visceral and somatic stimuli of the lower body, Azizs laboratory found similar brain activation to visceral (rectal) and somatic (anal) distention, although a greater activation of motor cortex by the somatic stimulus was observed.61 Traceys group used fMRI scanning of the brain to compare the cortical processing of aversive visceral (rectal distention) and somatic heat stimuli in 10 healthy control subjects.80 Similar to the ndings in the chest, the relative unpleasantness of the subjective experience of the visceral mechanical stimuli was higher than that of the somatic thermal stimuli. Controlling for unpleasantness, similar activations were found for the 2 stimulation types including insula, dACC, and secondary motor area. Visceral stimuli were associated with deactivation of the perigenual cingulate cortex (pACC); a nding also reported in somatic pain studies,81 with a relatively greater activation of the right anterior insula. Somatic (but not visceral) pain was associated with left dorsolateral PFC, a region concerned with cognitive processes. In a follow-up study,82 the authors compared brainstem responses to intensity-matched rectal or cutaneous abdominal electrical stimuli. Similar signicant activation associated with both stimuli was observed in several brainstem regions including the PAG, the PBN, the LCC, and the nucleus cuneiformis. However, 2 regions showed greater responses during the visceral pain condition: a signicantly greater activation of a region that the authors identied as the nucleus cuneiformis (perhaps signifying activity in the PBN described above) and a signicant correlation of the right PAG with anxiety ratings. The authors concluded from these ndings that the observed differences might represent a greater nocifensive response and a greater emotive salience of visceral pain.

December 2006

THE BRAIN-GUT AXIS

1933

Two studies have looked at differences in central processing of somatic and visceral experimental stimuli in IBS patients. These studies follow from a series of psychophysiological studies showing increased perception of visceral stimulation in IBS,83,84 but less consistent ndings regarding IBS sensitivity to noxious somatic stimuli. However, depending on the somatic pain stimulus used, different investigators have reported normal,85 88 reduced,89 and enhanced90 perceptual responses to somatic pain stimuli. One of the two imaging studies done to date comparing visceral and somatic stimuli in IBS used thermal pain90 and the other cutaneous pressure.89 Verne et al studied brain responses with fMRI to phasic aversive rectal distention and to tonic cutaneous heat applied to the foot in 9 IBS patients and in a group of healthy age- and sex-matched controls.90 Both types of nociceptive stimuli evoked greater neural activity in brain regions of patients compared with controls in dACC and insula, as well as in thalamus and somatosensory regions and in the PFC. Enhanced brain responses to both types of stimuli were observed within the same brain structures. The authors interpreted these ndings as supporting their hypothesis that visceral and cutaneous hypersensitivity in IBS patients is related to increased afferent processing in ascending pathways, rather than to altered cognitive and/or emotional modulation at higher brain levels. A somewhat different conclusion was reached from a study in female patients with IBS with (n 10) and without (n 10) a comorbid diagnosis of bromyalgia.91 Brain responses to somatic pressure and to aversive rectal distention were evaluated with H215O-PET. The somatic stimulus was perceived as less aversive than the visceral stimulus by the IBS patients, while IBSbromyalgia patients rated both stimuli as equally aversive. Group differences in regional brain activation were only observed within the dACC (not in the insula), where IBS patients showed a greater response to visceral stimuli and IBSbromyalgia patients showed a greater response to somatic stimuli. The authors concluded from their ndings that chronic stimulus-specic enhancement of dACC responses to sensory stimuli in both syndromes may be associated with cognitive enhancement of either visceral (IBS) or somatic (IBSbromyalgia) sensory input. The fact that no group differences were observed in primary sensory areas (thalamus, somatosensory cortex, insula) is consistent with the concept that afferent input that reaches the brain is not different between the 2 patient populations, while arousal and attentional mechanisms may differ. In summary, a growing number of brain imaging studies have addressed the question of how brain responses to somatic and visceral pain stimuli may differ, both in healthy control subjects, as well as in patients with IBS. Consistently, greater subjective affective rating of visceral pain stimuli (in terms of unpleasantness), and activation of the homeostatic afferent network by both types of stimuli has been observed in most studies. In contrast, a consistent difference in brain processing of visceral and somatic stimuli has not emerged. For example, evidence for the expected greater activation of limbic and paralimbic brain regions, in particular the dACC for visceral stimuli, or differences in arousal and antinociceptive mechanisms between visceral and somatic stimuli have not been reported. This lack of consistency may be due in large part to differences in study design (imaging modality, study paradigms, nature of stimuli used, previous exposure of subjects to similar stimuli,

sex of participants, etc.) and in the relatively small number of studies that have directly compared the 2 modalities. Based on the concept of homeostatic emotions presented earlier, one would expect that visceral stimuli of matched intensity may be associated with greater unpleasantness and motivational drive, a greater activation of dACC, and greater autonomic responses, compared with somatic stimuli.

Modulation of Homeostatic Afferent Processing Network by Cognitive and Emotional Stimuli

The brain has multiple ways to modulate the perception of afferent information and this modulation is inuenced by the environmental context, the emotional state of the individual (eg, fear, anxiety, or anger), cognitive factors (eg, expectation, attention), or memories of previous sensory events (conditioned responses) (Figure 2B). As outlined in the basic section of this review, these top-down inuences can modulate homeostatic afferent input to the central nervous system at multiple levels (Figure 2A). Considerable progress has been made both on a preclinical and, more recently, on a clinical level to identify brain regions, circuits, and mechanisms that play a role in the facilitation and inhibition of the subjective pain experience.46,92 Both clinical and preclinical studies support a role for right orbitofrontal and right ventrolateral PFC in mediating inhibition of emotional and pain responses (reviewed in Lieberman et al 200420) and for dorsolateral PFC in pain modulation.18,93

Emotional Modulation
Phillips et al94 studied the effects of emotional context on responses to nonpainful esophageal distention in 8 healthy subjects (7 males) on brain activation using fMRI. Activation within the right anterior insular cortex and bilateral dACC by the visceral stimulus was signicantly greater while simultaneously viewing fearful faces compared with neutral faces. In a second paradigm, subjective discomfort and brain responses in another 8 healthy male subjects during the same esophageal stimulus were studied while viewing faces with low, moderate, or high intensity of fear expression. During the high intensity fearful visual stimulus, signicantly greater discomfort, anxiety, and greater brain activation in the left dACC and bilateral anterior insular cortex was seen. These ndings clearly demonstrated the powerful effect of emotional context on the perceptual, emotional, and brain response to an innocuous visceral stimulus. Specically, they demonstrated clearly the modulatory role of emotion regulation circuits on the homeostatic afferent processing network.

Cognitive Modulation
Attention. Azizs group95 examined the modulatory role of attention on the brain responses to nonpainful visceral (esophageal) distention in 7 healthy volunteers (6 males). Brain responses to phasic visual and esophageal stimuli were presented simultaneously while subjects were asked to focus their attention on either the esophageal or the visual stimulus (selective attention) or both (divided attention). Selective attention on the esophageal stimulus was associated with activation of sensory (somatosensory cortex) and cognitive (dACC) networks, while selective attention on the visual stimulus activated the visual cortex. During the divided attention task, more brain regions in the sensory and cognitive domains were activated to

1934

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

Figure 4.

Central arousal networks and visceral hypersensitivity. Normalization of discomfort visual analogue scale ratings (left upper graph), and discomfort threshold (right upper graph) in 24 IBS patients undergoing experimental rectal balloon distention every 3 months over a 12-month period. Brain images obtained in a subset of 12 individuals during the initial and the fourth experimental session showed decreases in central arousal circuits involving the mid-cingulate cortex (MCC), amygdala (not shown), and dorsal pons (including LCC region). Similar decreases were observed during an expectation condition and during the actual distention. Reprinted with permission from Naliboff et al.101

process esophageal stimuli, in comparison to those processing visual stimuli. These ndings emphasize the importance of attentional processes in the modulation of sensory information from the body and the relative biological importance placed on visceral sensation, compared with other sensory modalities. Expectation. A variety of studies in the somatic pain literature have evaluated brain responses to the expectation of an aversive stimulus.59,75,81,96,97 These studies suggest that the brain can either up-regulate or down-regulate sensory and limbic brain regions based on previous experience, familiarity with the stimulus and expected intensity.81 Several early studies have shown some evidence of activation of the homeostatic afferent network during conditioned anticipation of a visceral stimulus.59,75 In a preliminary report, Berman et al studied the brain fMRI BOLD response to anticipated (cue condition) and delivered mild and moderate rectal distention in 12 healthy women and 14 female IBS patients.98 Distention increased activity in the homeostatic brain regions and decreased activity in the infragenual cingulate. As in previous studies comparing IBS patients and control subjects, the increases were more extensive in the IBS patients, with signicant differences in midcingulate and dorsal brainstem. During cued anticipation of distention, activity decreased in the insula, dACC, amygdala, and dorsal brainstem in healthy women, but not in IBS patients, consistent with a top-down modulation of homeostatic afferent networks by cortical regions. Three self-rated measures of negative affect during scanning were higher in IBS patients than healthy women (P .001), and the anticipatory BOLD decreases in bilateral dorsal brainstem, centered in the pontine LCC, were inversely correlated with all 3 measures. The amplitude of anticipatory decrease in the LCC was associated with greater activation by subsequent distention in right orbitofrontal cor-

tex and bilateral supragenual ACC regions previously implicated in cognitive coping. These ndings suggest that in healthy women, the brain decreases activity within the homeostatic brain matrix in expectation of a certain, inescapable pelvic pain stimulus. A failure to generate this down-regulation in IBS patients may be related to differences in cortically mediated coping styles, emotional factors, and linked arousal systems. Yaguez et al99 studied brain responses during different phases of visceral aversive conditioning in 8 healthy volunteers (5 males) using fMRI. The authors used a classical conditioning paradigm in which different colored circles were used as conditioned stimuli and were paired with painful esophageal distention (learning phase), airpuff to the wrist, or nothing. Brain responses during the learning phase (delivery of aversive esophageal distention) were seen in the homeostatic brain matrix and in somatosensory cortex. During the anticipation and extinction phase of the paradigm, brain activity resembled that seen during actual esophageal distention, including activation in insula and dACC. These ndings emphasize the importance of cognitive inuences, such as expectation and memory recall in top-down modulation of brain regions involved in the processing of homeostatic information from the body. Hypervigilance. Several lines of evidence indicate that IBS patients and other functional disorders have hypervigilance for symptom-relevant sensations.100 Repeated exposure to experimental visceral stimuli can lead to decreased hypervigilance and, therefore, discomfort. In a longitudinal study of IBS patients exposed to 6 sessions of rectal inations over a 1-year period, we examined regional cerebral blood ow to the inations and anticipation of inations using H215O-PET at the rst and last session.101 As shown in Figure 4, subjective ratings of the rectal inations normalized over the 12 months

December 2006

THE BRAIN-GUT AXIS

1935

of the study, while IBS symptom severity did not, indicating decreased vigilance independent of changes in perceived disease activity. In response to rectal distention, stable activation of regions of the homeostatic afferent network (including thalamus and anterior insula) was observed over the 12-month period, while activity in limbic, paralimbic, and pontine regions decreased. During the anticipation condition, there were significant decreases in dACC, amygdala, and dorsal brainstem (perhaps involving the LCC) activation at 12 months. One way to interpret these ndings is that brain regions processing the feeling and the motivational dimension of the homeostatic emotion were affected differentially by the habituation process: while insula activation remained constant, dACC activation progressively decreased. An analysis examining the covariation of these brain regions, as well as preliminary results from an effective connectivity modeling approach to the data,102 supported the hypothesis of changes in an arousal network including limbic, pontine, and cortical areas underlying the decreased perception seen over the multiple stimulation studies. In summary, brain imaging studies of cognitive and affective modulation of perceptual and brain responses to visceral stimuli strongly support the concept of a homeostatic afferent network in the brain. The fact that components of this network can be modulated differentially by top-down corticolimbic inuences has important implications for a better understanding of symptom generation/modulation in functional GI disorders.

Figure 5. Role of corticolimbic inhibitory network in pain modulation. Medial frontal cortical (FC) mediation of the impact of right lateral FC on the PAG. The dotted line indicates that the relationship between right lateral PFC [RLPF] and the PAG is indirect. Unstandardized coefcients with associated standard errors in parentheses are reported. The coefcients in brackets indicate the direct effect of RLFC on PAG prior to the inclusion of the mediating medial FC. ** indicates P .01. FC, frontal cortex. PAG, periaqueductal gray. RLFC, right lateral frontal cortex. Reprinted with permission from Mayer et al.109
tion.74,109,110 Two studies were performed by Wilder-Smith and coworkers74,110 to study the central correlates of heterotopic pain inhibition. In the rst study, they performed an fMRI study in 10 female patients with IBS (5 constipated-, 5 diarrheapredominant bowel habit) and 10 female healthy control subjects to test the hypothesis that IBS patients show abnormal activation of diffuse noxious inhibitory control (DNICs) systems in response to a noxious stimulus.74 DNIC activation can be quantied by the perceptual modulation of a painful stimulus (in this case noxious rectal balloon distention) by a secondary heterotypically applied nociceptive stimulus (in this case ice water immersion of the foot). They found that subjective pain ratings of rectal volume distention by the heterotypic cold pain stimulus was reduced in healthy controls but not in the IBS patient group, suggesting an inadequate activation of DNICs in the IBS patients. Following the heterotypic cold stimulus, a complex set of differences in response to rectal pain was found among the controls and the 2 IBS bowel habit based sub-groups. These included decreased activation in insula, thalamus, and PAG in the control group (perhaps reecting the DNIC process) that was absent in the IBS patients. In the second study, similar perceptual results were found and differences were also found between IBS and controls during expectation of rectal pain, without actual distention.110 More recent brain imaging studies in patient populations provide more direct support for alterations in cortico-limbicpontine pain modulation networks in IBS patients leading to visceral hypersensitivity. Mayer et al109 examined 3 groups of male subjects, ulcerative colitis patients with quiescent disease (n 9), patients with IBS (n 9), and healthy male controls (n 9), during actual and expected but undelivered rectal distentions using H215O-PET. This study found similar responses in all 3 groups in the homeostatic afferent network (anterior insula and dACC). However, IBS patients compared with both the ulcerative colitis and control groups showed consistently greater activation of limbic/paralimbic brain regions (amygdala, hypothalamus, ventral/rostral ACC, dorsomedial PFC) sugges-

Activation of Brain Networks Associated With Descending Pain Modulation

Since the beginning of the 20th century it has been known that the brain can tonically inhibit spinal cord excitability, thereby regulating the amount of peripheral sensory information reaching the central nervous system. More recent evidence has demonstrated the activity of both pain inhibitory and facilitatory mechanisms that can tonically and phasically regulate spinal cord excitability.44 46 While top-down tonic pain inhibitory modulation appears to predominate in healthy individuals during basal conditions, an up-regulation of descending pain facilitatory systems has been demonstrated in the maintenance of hyperalgesia in animal models of peripheral nerve injury.47 An alteration in the balance between inhibitory and facilitatory pain modulatory systems has been proposed as a possible mechanism underlying chronic pain syndromes such as bromyalgia103 and IBS.104,105 Zambreanu et al were the rst to demonstrate the activation of brainstem regions in the context of central sensitization in healthy human volunteers.106 Using 3T fMRI, they compared whole brain responses, including the brainstem, to punctuate mechanical stimulation in an area of secondary hyperalgesia (induced by heat/capsaicin sensitization model) or in a control area. They found greater activation during stimulation of the hyperalgesic region in several cortical regions, including posterior insula, ACC, and posterior cingulate cortex, as well as thalamus and pons. The brainstem activation was localized to the NCF (possibly involving PBN) and the PAG, brain regions that receive input from corticolimbic networks (including the rostral ACC), send projections to the rostroventral medulla and are part of a cortico-limbic-pontine pain modulation circuit107,108 (see also Figures 2 and 3). There is preliminary evidence to suggest that patients with IBS may also show abnormal activation of brain circuits involved in pain modula-

1936

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

tive of increased activation of pain facilitatory pathways. In addition, the results showed activation in the ulcerative colitis and control subjects, but not IBS patients, in the lateral frontal regions and a brain region including the PAG. A connectivity analysis using structural equation modeling supported these regions acting as part of a pain inhibition network that involves lateral and medial frontal inuences on the PAG (Figure 5). In summary, a small group of hypothesis-driven studies aimed at evaluating specic mechanisms of visceral pain modulation (eg, descending modulation) is beginning to demonstrate the activation of cortico-limbic-pontine networks, which may be the biological substrate of such endogenous modulation. The concept emerging from these studies is the ability of brain regions and networks involved in emotion generation (limbic circuits), as well as in emotion regulation (orbitofrontal, ventrolateral PFC) to modulate activity within the homeostatic afferent processing network.48,49 Differences in the activation of these networks by cognitive or somatic stimuli between IBS patients and control populations suggest a possible role of such differences in the pathophysiology of enhanced visceral perception in IBS.

Evidence for Altered Brain-Gut Interactions in Patients With Functional GI Disorders

Central correlates of visceral hypersensitivity. One important question in the pathophysiology of functional GI disorders is related to the nature and pathophysiology of the enhanced perceptual responses observed in a large number of experimental pain studies.83 A related question is the pathophysiology of alterations in autonomic reex responses to various gut stimuli. Although acute gut inammation in both human and animal experimental models is typically associated with peripheral and central sensitization (as well as important, time-dependent modulatory inuences from the brain in form of both facilitatory and inhibitory modulation), chronic visceral pain is likely to involve additional sensitization of spino-bulbospinal loops47 and of supraspinal mechanisms, as well as cognitive and emotional modulation of the chronic visceral pain experience related to coping mechanisms and symptom-related anxiety.111 Brain imaging in human subjects is one of the few experimental approaches which may be able to differentiate between these different pain modulation mechanisms (Figure 3). Such dissection of different pain modulation mechanisms and the identication of specic alterations in one or several of these mechanisms is the prerequisite for evaluating the effect of novel treatment approaches on the brain gut axis. Findings from the initial descriptive imaging studies51,58,59,112 suggested that patients showed similar areas of activation to controls67 but evidenced greater activation in some regions, especially in the dACC and perhaps in certain limbic regions, including the hypothalamus, infragenual cingulate cortex, and amygdala.59 Decreased activation in the dorsal pons (including the region of the PAG) was also reported in IBS patients,59 and these results gave rise to an initial hypothesis that patients might have increased affective and attentional responses to actual or anticipated visceral stimuli (hypervigilance), as well as potentially decreased descending pain inhibition.59 Sex-related differences. A series of observations demonstrate that women are more likely to suffer from IBS, develop the so-called post-infectious IBS, and develop comor-

bidities such as bromyalgia or interstitial cystitis.113 A variety of interrelated mechanisms have been proposed to explain these sex differences, including sex differences in emotion regulation, in central stress responsiveness, and in pain modulation systems.114,115 Berman et al reported the rst study of brain responses in 2 samples of a total of 30 IBS patients (13 females; 6 with constipation-predominant bowel habit) with H215O-PET in response to rectal distention.62 Despite similar subjective stimulus ratings by male and female patients, regional brain activations were stronger in males. In males, but not females, rectal distention was associated with activation of the 2 regions concerned with homeostatic emotions, the anterior insula, and dACC. Insula activation correlated most strongly with the objective intensity of the stimulus (rectal pressure), whereas dACC activation correlated most strongly with the affective response, eg, the subjective discomfort rating of the stimulus. Naliboff et al studied brain responses in 42 (23 females) nonconstipated IBS patients to a visceral (rectal) stimulus of moderate intensity and during expectation of such distention using H215O-PET.75 In response to the visceral stimulus, both male and female patients showed activation of the homeostatic brain matrix in addition to PFC and brainstem regions. Female patients showed greater activation in the dACC, as well as in limbic (amygdala) and paralimbic regions (ventromedial PFC, infragenual cingulate cortex), whereas male patients showed greater activation of the midposterior insula, dorsolateral PFC, and dorsal pons. Similar sex-related differences were observed during the expectation condition, without any visceral stimulus. This study replicated the nding from the earlier study showing greater activation by male IBS patients of the insula. The ndings also suggested that female patients in response to a pelvic aversive stimulus show greater responses of limbic and paralimbic regions, while male patients show greater activation of regions belonging to a corticolimbic pain inhibition system. Preliminary results from an effective connectivity analysis of the same data set suggest that the observed differences are related to differences in the connectivity within an arousal circuit, involving the amygdala, dorsal brainstem, and dACC.116 Based on the neuroimaging ndings summarized previously, we are proposing a conceptual model for central nervous system abnormalities in IBS patients. This model is based on the general concept that IBS patients are a heterogeneous group of patients that share alterations in the perceptual and reex responses to homeostatic afferent signals from the GI tract. The biological substrate of this model is the homeostatic afferent processing network (in particular the insula and dACC), with a range of modulatory inuences on the activity of this network, ranging from bottom-up sensitized input (central sensitization, increased spinobulbo-spinal pain amplication loops) or sensitized topdown modulation (arousal and emotion generation networks), or ineffective top-down inhibition (corticolimbic networks) (summarized in Figure 5). The model is consistent with the frequent overlap of visceral, somatic, and emotional functional syndromes, and with the prominent role of affect and mood in clinical presentations. Different aspects of this model are amenable to validated rigorous study paradigms using neuroimaging techniques and should yield conclusive

December 2006

THE BRAIN-GUT AXIS

1937

results in adequately powered studies using well-dened study populations.

Modulation of Brain Responses in IBS by Therapeutic Interventions

The application of functional brain imaging techniques to drug discovery and candidate compound evaluation (pharmacological fMRI) has great potential for the eld of functional GI disorders.117 In addition to providing information about structure and function of the brain, this approach may improve the translation from animal models of these disorders118 to the human syndrome, as well as expedite the decision making process if a candidate compound is likely to work in human patients. The assessment of symptoms and their improvement with therapeutic interventions in patients with functional GI disorders depends currently on the subjective patient reports obtained in lengthy and expensive phase II and III clinical trials. The need for large patient numbers in these studies, requiring many participating centers, sometimes in different parts of the world, further increases the variability of subjective data. The use of functional imaging techniques, in particular fMRI has the potential to provide sensitive and specic biomarkers relevant to the symptom complex of functional bowel syndromes, such as primary visceral afferent hypersensitivity, increased activity in central arousal systems, compromised endogenous pain inhibition systems and other pain amplication mechanisms. Additional biomarkers include activity of and connectivity within central autonomic networks. The ability to sample a specic neurobiological substrate (eg, the amygdala or the insula) hundreds of times in the same subject, greatly reduces the intra- and inter-individual variability of the measure. Thus, neuroimaging-based evaluation of drug effects in small samples of patients has the potential to identify signicant drug effects on a specic biomarker of the disorder, while much larger patient numbers are required to detect signicant effects on subjective symptoms. Since fMRI can be applied to animal models as well, identication of drug effects in preclinical studies may have predictive validity for nding such effects in human studies as well. Several studies outside of the eld of Gastroenterology have reported the ability of functional brain imaging to detect changes in brain activation induced by various therapeutic interventions, ranging from hypnosis,19 placebo,20,96,97 cognitive behavioral therapy,119 antidepressants,120 anxiolytics,121,122 opioids123 and gabapentin.124 These results have demonstrated that several of these interventions affect activity within the homeostatic afferent processing network, as well as specic circuits in the brain that play a role in arousal and in endogenous pain modulation, such as amygdala and pontine regions. A particularly interesting example supporting the concept of dACC as a region involved in affect and motivational responses to painful stimuli is the biofeedback training of subjects to increase or decrease the activation of the dACC, using real time fMRI.125 While healthy subjects could increase or decrease pain perception to an experimental stimulus based on their volitional control of dACC activity, chronic pain patients were able to decrease their spontaneous pain by decreasing dACC activity. Despite the lack of consensus regarding brain responses to rectal stimuli in healthy controls and group differences between IBS patients and control subjects discussed above, several studies have been reported using functional brain imaging to identify

changes in cerebral activation associated with various treatment modalities, including pharmacological126 128 and non-pharmacological treatments.20,129 In contrast to the emerging publications in the somatic pain and psychiatric literature,121,122,124,130 only a few of the reported studies were of sufcient quality (statistical power, blinding, homogeneous study populations) to allow any conclusions from the results. Pharmacological interventions. Morgan et al reported an effect of low-dose amitriptyline treatment (compared with placebo) on rectal distentioninduced brain activation combined with a psychological stressor using fMRI.126 Although the drug had no effect on symptoms, or distentions alone, decreased activation in the rostral ACC and the left posterior parietal cortex was observed during distention when associated with the psychological stressor. Berman et al reported a double-blind, randomized, placebo-controlled study in 49 IBS patients (26 women; nonconstipated) who underwent H215O-PET scanning before and after a 3-week course of the 5-HT3 receptor antagonist, alosetron.127 Active treatment was associated with reduced blood ow in limbic and paralimbic regions, including the amygdala and the ventral striatum, ventromedial PFC, and a pontine region, but not with signicant changes in the homeostatic afferent network (insula, dACC, thalamus). Signicant reductions were only seen at baseline and during the expectation condition, but not during the rectal distention, suggesting that the main effect of the drug was on brain networks engaged during expectation, rather than on visceral afferent pathways. Supporting this interpretation was the nding that IBS symptom improvement was correlated with regional cerebral blood ow in the amygdala, ventral striatum, and dorsal pons. In a preliminary study, Kilpatrick et al used fMRI to characterize the effect of tegaserod (vs placebo) on brain responses to rectal distention in a crossover study of 10 female IBS patients.131 Effective connectivity analysis revealed a distention-activated network of regions associated with the medial thalamus that functioned differently during tegaserod and placebo treatment. This treatment-dependent medial thalamic network included the homeostatic afferent network (insula, dACC), as well as a cortico-limbic network (amygdala, ventral ACC, and prefrontal cortices). Thus, despite the small sample size, the study results suggest drug-induced differences in the functioning of a spinal-thalamic-cortico-limbic circuit operating during experience of noxious visceral stimuli. Behavioral interventions. Drossman et al reported a case study of a severe female IBS patient with signicant psychological distress, a history of sexual abuse, and severe functional GI symptoms.132 Evaluation during symptom are showed high subjective ratings of GI and psychological symptoms, a low rectal discomfort threshold, and distention-induced activation of multiple brain regions, including the dACC. Symptom improvement was associated with normalization of the perceptual hypersensitivity to rectal balloon distention and with a signicant decrease in distention-induced activation of the dACC, the somatosensory cortex, and a PFC region.133 Lackner et al studied 6 women with IBS using H2O15-PET before and after a brief course of cognitive behavioral therapy.134 Comparing pretreatment resting scans (without rectal distention) with posttreatment scans, a reduction in regional cerebral blood ow was found in the parahippocampal gyrus and the ventral portion of the ACC. These brain changes were associated with signicant improvement in GI symptoms and

1938

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

psychological functioning. Reduced blood ow in the left pons was correlated with posttreatment anxiety ratings. In summary, the information gained from a small number of published neuroimaging studies of brain activity associated with treatment responses in IBS patients has to be considered as preliminary. In the absence of control conditions, any observed change in the response of the brain to a rectal stimulus may simply be related to a habituation to the experimental stimulus, rather than a drug effect. The nding of selective effects of alosetron treatment on limbic, but not primary pain regions, and the correlation of these limbic effects with IBS symptom ratings demonstrates the potential strength of this technique to understanding the action of new IBS treatments. Well-designed treatment studies, with adequate sample size, homogeneous study populations, and reproducible study paradigms are needed to conrm the validity of this approach to monitor treatment effects, and predict possible clinical outcomes. However, the potential benet of the pharmacological fMRI approach to drug discovery and evaluation is considerable.135 Candidate compounds aimed at ascending and descending modulatory inuences on the homeostatic afferent processing network can be evaluated in their effectiveness and pharmacokinetics on animal models, and on their clinical relevance in small early phase II studies. For example, candidate IBS drugs currently being evaluated with pharmacological fMRI include antagonists for the corticotropin releasing factor (CRF)1 receptor and the NK1 receptor.

Table 1. Possible Future Clinical Applications of fMRI in Brain Gut Disorders

Application of fMRI for understanding disease Characterization of mechanisms of disease109 Markers of disease-related traits (eg, arousal, hypervigilance)101,140 Endophenotypes for genetic characterization of disease136 In vivo assay for functional polymorphism141 Applications of fMRI for the discovery and development of new therapies117,135 Relating molecular targets to behaviors Enrichment of study populations with treatment responders142 Differentiating strong placebo responders97 Pharmacodynamic markers124 Pharmacokinetic markers123 More sensitive measures of treatment response127,143 Modied from Matthews et al.139

terms of activation and connectivity), and underlying biology and genetics on an unprecedented scale. References
1. Craig AD. An ascending general homeostatic afferent pathway originating in lamina I. Prog Brain Res 1996;107:225242. 2. Craig AD. How do you feel? Interoception: The sense of the physiological condition of the body. Nat Rev Neurosci 2002;3: 655 666. 3. Damasio AR. The feeling of what happens: body and emotion in the making of consciousness. 1st ed. New York, NY: Harcourt Brace, 1999. 4. Rolls ET. The brain and emotion. Oxford: Oxford University Press, 1999. 5. Mayer EA, Saper CB, eds. The biological basis for mind body interactions. Amsterdam: Elsevier, 2000. 6. Mayer EA. The neurobiology of stress and gastrointestinal disease. Gut 2000;47:861 869. 7. Faris PL, Kim SW, Meller WH, Goodale RL, Hofbauer RD, Oakman SA, Howard LA, Stevens ER, Eckert ED, Hartman BK. Effect of ondansetron, a 5-HT3 receptor antagonist, on the dynamic association between bulimic behaviors and pain thresholds. Pain 1998;77:297303. 8. Whitehead WE, Palsson O, Jones KR. Systemic review of the comorbidity of irritable bowel syndrome with other disorders: what are the causes and implications? Gastroenterology 2002; 122:1140 1156. 9. Wessely S, Nimnuan C, Sharpe M. Functional somatic syndromes: one or many? Lancet 1999;354:936 939. 10. Sengupta JN, Gebhart GF. Gastrointestinal afferent bers and sensation. In: Johnson LR, ed. Physiology of the gastrointestinal tract. 3rd ed. New York: Raven, 1994:483519. 11. Wilson LB, Andrew D, Craig AD. Activation of spinobulbar lamina I neurons by static muscle contraction. J Neurophysiol 2002; 87:16411645. 12. Andrew D, Craig AD. Spinothalamic lamina I neurons selectively sensitive to histamine: a central neural pathway for itch. Nat Neurosci 2001;4:7277. 13. Craig AD, Andrew D. Responses of spinothalamic lamina I neurons to repeated brief contact heat stimulation in the cat. J Neurophysiol 2002;87:19021914. 14. Andrew D, Craig AD. Quantitative responses of spinothalamic lamina I neurones to graded mechanical stimulation in the cat. J Physiol 2002;545:913931. 15. James W. Principles of psychology. New York: Holt, 1890. 16. Darwin C. The expression of the emotions in man and animals. Chicago, IL: The University of Chicago Press, 1965.

Summary and Future Directions

The concept of homeostatic emotions with a sensory feeling dimension (processed in the anterior insula) and an affective/motivational dimension (processed in the dACC) provides a general framework to understand the interactions between peripheral afferent signals arising from the gut, and centrally mediated inuences (including psychological factors and emotions) on the conscious perception, affective response, and associated autonomic responses to such stimuli. The deconstruction of symptoms into distinct contributions of specic brain networks mediating a variety of cognitive, emotional, and motivational inuences on the basic homeostatic afferent processing network (Figure 3) may help to understand the pathophysiology of a variety of chronic disorders, thought to involve brain-gut interactions, including but not limited to functional GI disorders, eating disorders, and neuroimmune interactions in inammatory bowel disease. A whole range of chronic disorders characterized by physical or emotional discomfort and pain (including functional visceral and somatic disorders, as well as disorders of mood and affect) can be conceptualized as disorders of homeostatic emotions. The ability to study a neurobiological substrate with different imaging modalities, rather than relying on highly variable subjective symptoms, makes it possible to obtain insights about the role of genetic factors,136 receptor physiology,137 and therapeutic interventions135 on distinct brain networks involved in afferent processing and modulation. Meaningful results from such studies can be obtained from much smaller samples of subjects, compared with epidemiological or traditional pharmacological studies. A list of possible future clinical application of fMRI is shown in Table 1. Multimodal imaging approaches in the same subject enable investigators to correlate structure, function (in

December 2006

THE BRAIN-GUT AXIS

1939

17. Vogt BA. Pain and emotion interactions in subregions of the cingulate gyrus. Nat Rev Neurosci 2005;6:533544. 18. Lorenz J, Minoshima S, Casey KL. Keeping pain out of mind: the role of the dorsolateral prefrontal cortex in pain modulation. Brain 2003;126:1079 1091. 19. Rainville P, Duncan GH, Price DD, Carrier B, Bushnell MC. Pain affect encoded in human anterior cingulate but not somatosensory cortex. Science 1997;277:968 971. 20. Lieberman MD, Jarcho JM, Berman S, Naliboff BD, Suyenobu BY, Mandelkern M, Mayer EA. The neural correlates of placebo effects: a disruption account. Neuroimage 2004;22:447 455. 21. Schmidt RF. Leben ohne schmerz - die wirklichkeit von morgan? In: Kerger H, ed. Schmerz und Schmerztherapie. Kassel: Akad frzt Fortibildung u Weiterbildung Bad Neuheim, 1984. 22. Ness TJ, Gebhart GF. Colorectal distention as a noxious visceral stimulus: physiological and pharmacological characterization of pseudoaffective reexes in the rat. Brain Res 1988;450:153 169. 23. Altman J, Bayer SA. The development of the rat spinal cord. Adv Anat Embryol Cell Biol 1984;85:1164. 24. Jnig W. The integrative action of the autonomic nervous system: neurobiology of homeostasis. New York, NY: Cambridge University Press, 2006. 25. Cervero F. In: Belmonte C, Cervero F, eds. Neurobiology of nociceptors. Oxford University Press, 1996:220 242. 26. Adelson DW, Wei JY, Kruger L. Warm-sensitive afferent splanchnic C-ber units in vitro. J Neurophysiol 1997;77:2989 3002. 27. Grundy D, Al-Chaer ED, Aziz Q, Collins SM, Ke M, Tach Y, Wood JD. Fundamentals of neurogastroenterology: basic science. Gastroenterology 2006;130:13911411. 28. Gershon MD. Nerves, reexes, and the enteric nervous system: pathogenesis of the irritable bowel syndrome. J Clin Gastroenterol 2005;39:S184 S193. 29. Barbara G, Stanghellini V, De Giorgio R, Cremon C, Cottrell GS, Santini D, Pasquinelli G, Morselli-Labate AM, Grady EF, Bunnett NW, Collins SM, Corinaldesi R. Activated mast cells in proximity to colonic nerves correlate with abdominal pain in irritable bowel syndrome. Gastroenterology 2004;126:693702. 30. Brierley SM, Jones RC3, Gebhart GF, Blackshaw LA. Splanchnic and pelvic mechanosensory afferents signal different qualities of colonic stimuli in mice. Gastroenterology 2004;127:166 178. 31. Lynn P, Zagorodnyuk V, Hennig G, Costa M, Brookes S. Mechanical activation of rectal intraganglionic laminar endings in the guinea pig distal gut. J Physiol 2005;564:589 601. 32. Rosas-Arellano MP, Solano-Flores LP, Ciriello J. c-Fos induction in spinal cord neurons after renal arterial or venous occlusion. Am J Physiol 1999;276:R120 R127. 33. Schmelz M, Schmid R, Handwerker HO, Torebjork HE. Encoding of burning pain from capsaicin-treated human skin in two categories of unmyelinated nerve bres. Brain 2000;123:560 571. 34. Konturek SJ, Konturek JW, Pawlik T, Brzozowski T. Brain-gut axis and its role in the control of food intake. J Physiol Pharmacol 2004;55:137154. 35. Reidelberger RD, Hernandez J, Fritzsch B, Hulce M. Abdominal vagal mediation of the satiety effects of CCK in rats. Am J Physiol Regul Integr Comp Physiol 2004;286:R1005R1012. 36. Sato A, Schmidt RF. Somatosympathetic reexes: afferent bers, central pathways, discharge characteristics. Physiol Rev 1973;53:916 947. 37. Diesel DA, Tucker A, Robertshaw D. Cold-induced changes in breathing pattern as a strategy to reduce respiratory heat loss. J Appl Physiol 1990;69:1946 1952. 38. Mitchell JH, Schmidt RF. Cardiovascular reex control by afferent bers from skeletal muscle receptors. In: Shepard JT, Abboud FM, eds. The handbook of physiology: section 2, The

39.

40.

41.

42. 43. 44. 45. 46. 47.

48. 49.

50. 51.

52.

53.

54.

55.

56.

57.

58.

cardiovascular system. Bethesda, MD: American Physiological Society, 1984:623 658. Johansen JP, Fields HL, Manning BH. The affective component of pain in rodents: direct evidence for a contribution of the anterior cingulate cortex. Proc Natl Acad Sci U S A 2001;98: 8077 8082. Critchley HD, Coreld DR, Chandler MP, Mathias CJ, Dolan RJ. Cerebral correlates of autonomic cardiovascular arousal: a functional neuroimaging investigation in humans. J Physiol 2000; 523:259 270. Critchley HD, Mathias CJ, Josephs O, ODoherty J, Zanini S, Dewar BK, Cipolotti L, Shallice T, Dolan RJ. Human cingulate cortex and autonomic control: converging neuroimaging and clinical evidence. Brain 2003;126:2139 2152. Critchley HD. The human cortex responds to an interoceptive challenge. Proc Natl Acad Sci U S A 2004;101:6333 6334. Craig AD. Forebrain emotional asymmetry: a neuroanatomical basis? Trends Cogn Sci 2005;9:566 571. Gebhart GF. Descending modulation of pain. Neurosci Biobehavi Rev 2004;27:729 737. Porreca F, Ossipov MH, Gebhart GF. Chronic pain and medullary descending facilitation. Trends Neurosci 2002;25:319 325. Tracey I, Dunckley P. Importance of anti- and pro-nociceptive mechanisms in human disease. Gut 2004;53:15531555. Suzuki R, Morcuende S, Webber M, Hunt SP, Dickenson AH. Supercial NK1-expressing neurons control spinal excitability through activation of descending pathways. Nat Neurosci 2002; 5:1319 1326. Petrovic P, Ingvar M. Imaging cognitive modulation of pain processing. Pain 2002;95:15. Villemure C, Bushnell MC. Cognitive modulation of pain: how do attention and emotion inuence pain processing? Pain 2002; 95:195199. Price DD. Psychological and neural mechanisms of the affective dimension of pain. Science 2000;288:1769 1772. Silverman DH, Munakata JA, Ennes H, Mandelkern MA, Hoh CK, Mayer EA. Regional cerebral activity in normal and pathological perception of visceral pain. Gastroenterology 1997;112:64 72. Kern MK, Birn RM, Jaradeh S, Jesmanowicz A, Cox RW, Hyde JS, Shaker R. Identication and characterization of cerebral cortical response to esophageal mucosal acid exposure and distention. Gastroenterology 1998;115:13531362. Kern MK, Arndorfer RC, Hyde JS, Shaker R. Cerebral cortical representation of external anal sphincter contraction: effect of effort. Am J Physiol Gastrointest Liver Physiol 2004;286:G304 G311. Binkofski F, Schnitzler A, Enck P, Frieling T, Posse S, Seitz RJ, Freund HJ. Somatic and limbic cortex activation in esophageal distention: a functional magnetic resonance imaging study. Ann Neurol 1998;44:811 815. Aziz Q, Andersson JL, Valind S, Sundin A, Hamdy S, Jones AK, Foster ER, Langstrom B, Thompson DG. Identication of human brain loci processing esophageal sensation using positron emission tomography. Gastroenterology 1997;113:50 59. Aziz Q, Thompson DG, Ng VWK, Hamdy S, Sarkar S, Brammer MJ, Bullmore ET, Hobson A, Tracey I, Gregory L, Simmons A, Williams SCR. Cortical processing of human somatic and visceral sensation. J Neurosci 2000;20:26572663. Rosen SD, Paulesu E, Frith CD, Frackowiak RSJ, Davies GJ, Jones T. Central nervous pathways mediating angina pectoris. Lancet 1994;344:147150. Mertz H, Morgan V, Tanner G, Pickens D, Price R, Shyr Y, Kessler R. Regional cerebral activation in irritable bowel syndrome and control subjects with painful and nonpainful rectal distention. Gastroenterology 2000;118:842 848.

1940

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

59. Naliboff BD, Derbyshire SWG, Munakata J, Berman S, Mandelkern M, Chang L, Mayer EA. Cerebral activation in irritable bowel syndrome patients and control subjects during rectosigmoid stimulation. Psychosom Med 2001;63:365375. 60. Ladabaum U, Minoshima S, Hasler WL, Cross D, Chey WD, Owyang C. Gastric distention correlates with activation of multiple cortical and subcortical regions. Gastroenterology 2001; 120:369 376. 61. Hobday DI, Aziz Q, Thacker N, Hollander I, Jackson A, Thompson DG. Study of the cortical processing of ano-rectal sensation using functional MRI. Brain 2001;124:361368. 62. Berman S, Munakata J, Naliboff B, Chang L, Mandelkern M, Silverman DH, Kovalik E, Mayer EA. Gender differences in regional brain response to visceral pressure in IBS patients. Eur J Pain 2000;4:157172. 63. Baciu MV, Bonaz BL, Papillon E, Bost RA, Le Bas JF, Fournet J, Segebarth CM. Central processing of rectal pain: a functional MR imaging study. Am J Neuroradiol 1999;20:1920 1924. 64. Lotze M, Wietek B, Birbaumer N, Ehrhardt J, Grodd W, Enck P. Cerebral activation during anal and rectal stimulation. Neuroimage 2001;14:10271034. 65. Bernstein CN, Frankenstein UN, Rawsthorne P, Pitz M, Summers R, McIntyre MC. Cortical mapping of visceral pain in patients with GI disorders using functional magnetic resonance imaging. Am J Gastroenterol 2002;97:319 327. 66. Kern MK, Jaradeh S, Arndorfer RC, Jesmanowicz J, Hyde J, Shaker R. Gender differences in cortical representation of rectal distention in healthy humans. Am J Physiol Gastrointest Liver Physiol 2001;281:G1512G1523. 67. Derbyshire SW. A systematic review of neuroimaging data during visceral stimulation. Am J Gastroenterol 2003;98:1220. 68. Peyron R, Laurent B, Garca-Larrea L. Functional imaging of brain responses to pain. A review and meta-analysis. Neurophysiologie Clinique 2000;30:263288. 69. Derbyshire SWG. Metareview of functional imaging studies with noxious stimuli reveals reproducible patterns of central activation. J Pain 2001;2:655. 70. Apkarian AV, Bushnell MC, Treede RD, Zubieta JK. Human brain mechanisms of pain perception and regulation in health and disease. Eur J Pain 2005;9:463 484. 71. Berman SM, Naliboff BD, Suyenobu B, Labus JS, Stains J, Bueller JA, Ruby K, Mayer EA. Sex differences in regional brain response to aversive pelvic visceral stimuli. Am J Physiol Regul Integr Comp Physiol 2006;291:R268 R276. 72. Toogood JA, Barr AM, Stevens TK, Gati JS, Menon RS, Martin RE. Discrete functional contributions of cerebral cortical foci in voluntary swallowing: A functional magnetic resonance imaging (fMRI) Go, No-Go study. Exp Brain Res 2005;161:8190. 73. Vandenbergh J, Dupont P, Fischler B, Bormans G, Persoons P, Janssens J, Tack J. Regional brain activation during proximal stomach distention in humans: a positron emission tomography study. Gastroenterology 2005;128:564 573. 74. Wilder-Smith CH, Schindler D, Lovblad K, Redmond SM, Nirkko A. Brain functional magnetic resonance imaging of rectal pain and activation of endogenous inhibitory mechanisms in irritable bowel syndrome patient subgroups and healthy controls. Gut 2004;53:15951601. 75. Naliboff BD, Berman S, Chang L, Derbyshire SWG, Suyenobu B, Vogt BA, Mandelkern M, Mayer EA. Sex-related differences in IBS patients: central processing of visceral stimuli. Gastroenterology 2003;124:1738 1747. 76. Hobson AR, Furlong PL, Worthen SF, Hillebrand A, Barnes GR, Singh KD, Aziz Q. Real-time imaging of human cortical activity evoked by painful esophageal stimulation. Gastroenterology 2005;128:610 619.

77. Strigo IA, Bushnell MC, Boivin M, Duncan GH. Psychosocial analysis of visceral and cutaneous pain in human subjects. Pain 2002;97:235246. 78. Strigo IA, Duncan GH, Boivin M, Bushnell MC. Differentiation of visceral and cutaneous pain in the human brain. J Neurophysiol 2003;89:3294 3303. 79. Strigo IA, Albanese MC, Bushnell MC, Duncan GH. Visceral and cutaneous pain representation in parasylvian cortex. Neurosci Lett 2005;384:54 59. 80. Dunckley P, Wise RG, Aziz Q, Painter D, Brooks J, Tracey I, Chang L. Cortical processing of visceral and somatic stimulation: differentiating pain intensity from unpleasantness. Neuroscience 2005;133:533542. 81. Porro CA, Cettolo V, Francescato MP, Baraldi P. Functional activity mapping of the mesial hemispheric wall during anticipation of pain. Neuroimage 2003;19:1738 1747. 82. Dunckley P, Wise RG, Fairhurst M, Hobden P, Aziz Q, Chang L, Tracey I. A comparison of visceral and somatic pain processing in the human brainstem using functional magnetic resonance imaging. J Neurosci 2005;25:73337341. 83. Mertz H. Review article: visceral hypersensitivity. Aliment Pharmacol Ther 2003;17:623 633. 84. Bouin M, Plourde V, Boivin M, Riberdy M, Lupien F, Laganiere M, Verrier P, Poitras P. Rectal distention testing in patients with irritable bowel syndrome: sensitivity, specicity, and predictive values of pain sensory thresholds. Gastroenterology 2002;122: 17711777. 85. Cook IJ, Van Eeden A, Collins SM. Patients with irritable bowel syndrome have greater pain tolerance than normal subjects. Gastroenterology 1987;93:727733. 86. Whitehead WE, Holtkotter B, Enck P, Hoelzl R, Holmes KD, Anthony J, Shabsin HS, Schuster MM. Tolerance for rectosigmoid distention in irritable bowel syndrome. Gastroenterology 1990;98:11871192. 87. Accarino AM, Azpiroz F, Malagelada JR. Selective dysfunction of mechanosensitive intestinal afferents in irritable bowel syndrome. Gastroenterology 1995;108:636 643. 88. Zighelboim J, Talley NJ, Phillips SF, Harmsen WS, Zinsmeister AR. Visceral perception in irritable bowel syndrome. Dig Dis Sci 1995;40:819 827. 89. Chang L, Mayer EA, Johnson T, FitzGerald L, Naliboff B. Differences in somatic perception in female patients with irritable bowel syndrome with and without bromyalgia. Pain 2000;84: 297307. 90. Verne GN, Robinson ME, Price DD. Hypersensitivity to visceral and cutaneous pain in the irritable bowel syndrome. Pain 2001; 93:714. 91. Chang L, Berman S, Mayer EA, Suyenobu B, Derbyshire S, Naliboff B, Vogt B, FitzGerald L, Mandelkern MA. Brain responses to visceral and somatic stimuli in patients with irritable bowel syndrome with and without bromyalgia. Am J Gastroenterol 2003;98:1354 1361. 92. Tracey I. Nociceptive processing in the human brain. Curr Opin Neurobiol 2005;15:478 487. 93. Apkarian AV, Sosa Y, Sonty S, Levy RM, Harden RN, Parrish TB, Gitelman DR. Chronic back pain is associated with decreased prefrontal and thalamic gray matter density. J Neurosci 2004; 24:10410 10415. 94. Phillips ML, Gregory LJ, Cullen S, Coen S, Ng V, Andrew C, Giampietro V, Bullmore E, Zelaya F, Amaro E, Thompson DG, Hobson AR, Williams SC, Brammer M, Aziz Q. The effect of negative emotional context on neural and behavioural responses to oesophageal stimulation. Brain 2003;126:669 684. 95. Gregory LJ, Yaguez L, Williams SC, Altmann C, Coen SJ, Ng V, Brammer MJ, Thompson DG, Aziz Q. Cognitive modulation of the cerebral processing of human oesophageal sensation using

December 2006

THE BRAIN-GUT AXIS

1941

96.

97.

98.

99.

100. 101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111. 112. 113.

functional magnetic resonance imaging. Gut 2003;52: 16711677. Petrovic P, Kalso E, Petersson KM, Ingvar M. Placebo and opioid analgesiaimaging a shared neuronal network. Science 2002; 295:17371740. Wager TD, Rilling JK, Smith EE, Sokolik A, Casey KL, Davidson RJ, Kosslyn SM, Rose RM, Cohen JD. Placebo-induced changes in fMRI in the anticipation and experience of pain. Science 2004;303:11621167. Berman SM, Naliboff BD, Suyenobu B, Chang L, Bueller J, Ruby K, Minnema M, Stains J, Mayer EA. Abnormal CNS response to anticipation of visceral distention in female patients with irritable bowel syndrome (IBS) an FMRI study. Gastroenterology 2006;130:A-78. Yaguez L, Coen S, Gregory LJ, Amaro EJr, Altman C, Brammer MJ, Bullmore ET, Williams SC, Aziz Q. Brain response to visceral aversive conditioning: a functional magnetic resonance imaging study. Gastroenterology 2005;128:1819 1829. Mayer EA, Craske MG, Naliboff BD. Depression, anxiety and the gastrointestinal system. J Clin Psychiatry 2001;62:28 36. Naliboff BD, Berman S, Suyenobu B, Labus JS, Chang L, Stains J, Mandelkern MA, Mayer EA. Longitudinal change in perceptual and brain activation response to visceral stimuli in irritable bowel syndrome patients. Gastroenterology 2006;131: 352365. Labus JS, Mayer EA, Berman SM, Suyenobu B, Chang L, Naliboff BD. Identication of brain networks associated with visceral perception and hypervigilance in patients with irritable bowel syndrome (IBS). Gastroenterology 2006;130:A-78. Clauw DJ, Crofford LJ. Chronic widespread pain and bromyalgia: what we know, and what we need to know. Best Pract Res Clin Rheumatol 2003;17:685701. Munakata J, Naliboff B, Harraf F, Kodner A, Lembo T, Chang L, Silverman DH, Mayer EA. Repetitive sigmoid stimulation induces rectal hyperalgesia in patients with irritable bowel syndrome. Gastroenterology 1997;112:55 63. Chang L, Naliboff BD, Labus JS, Schmulson M, Lee OY, Olivas TI, Stains J, Mayer EA. Effect of sex on perception of rectosigmoid stimuli in irritable bowel syndrome. Am J Physiol Regul Integr Comp Physiol 2006;291:R277R284. Zambreanu L, Wise RG, Brooks JC, Iannetti GD, Tracey I. A role for the brainstem in central sensitisation in humans. Evidence from functional magnetic resonance imaging. Pain 2005;114: 397 407. Petrovic P, Petersson KM, Hansson P, Ingvar M. Brainstem involvement in the initial response to pain. Neuroimage 2004; 22:9951005. Petrovic P, Dietrich T, Fransson P, Andersson J, Carlsson K, Ingvar M. Placebo in emotional processinginduced expectations of anxiety relief activate a generalized modulatory network. Neuron 2005;46:957969. Mayer EA, Berman S, Suyenobu B, Labus J, Mandelkern MA, Naliboff BD, Chang L. Differences in brain responses to visceral pain between patients with irritable bowel syndrome and ulcerative colitis. Pain 2005;115:398 409. Song GH, Venkatraman V, Ho KY, Chee MW, Yeoh KG, WilderSmith CH. Cortical effects of anticipation and endogenous modulation of visceral pain assessed by functional brain MRI in irritable bowel syndrome patients and healthy controls. Pain 2006;July 15 [Epub ahead of print] Mayer EA, Gebhart GF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107:271293. Kern MK, Shaker R. Cerebral cortical registration of subliminal visceral stimulation. Gastroenterology 2002;122:290 298. Mayer EA, Berman S, Chang L, Naliboff BD. Sex-based differences in gastrointestinal pain. Eur J Pain 2004;8:451 463.

114. Naliboff BD, Heitkemper MM, Chang L, Mayer EA. Sex and gender in irritable bowel syndrome. In: Fillingim RB, ed. Sex, gender, and pain. Seattle: IASP Press, 2000:327353. 115. Chang L, Heitkemper MM. Gender differences in irritable bowel syndrome. Gastroenterology 2002;123:1686 1701. 116. Labus J, Mayer EA, Berman S, Suyenobu B, Chang L, Naliboff BD. Sex-specic differences in a brain network functioning during anticipation of rectal discomfort in irritable bowel syndrome patients (IBS). Gastroenterology 2006;130:A-93. 117. Schweinhardt P, Bountra C, Tracey I. Pharmacological fMRI in the development of new analgesic compounds. NMR Biomed 2006;19:702711. 118. Mayer EA, Collins SM. Evolving pathophysiologic models of functional gastrointestinal disorders. Gastroenterology 2002; 122:20322048. 119. Goldapple K, Segal Z, Garson C, Lau M, Bieling P, Kennedy S, Mayberg H. Modulation of cortical-limbic pathways in major depression: treatment-specic effects of cognitive behavior therapy. Arch Gen Psychiatry 2004;61:34 41. 120. Fu CH, Williams SC, Cleare AJ, Brammer MJ, Walsh ND, Kim J, Andrew CM, Pich EM, Williams PM, Reed LJ, Mitterschiffthaler MT, Suckling J, Bullmore ET. Attenuation of the neural response to sad faces in major depression by antidepressant treatment: a prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry 2004;61:877 889. 121. Paulus MP, Feinstein JS, Castillo G, Simmons AN, Stein MB. Dose-dependent decrease of activation in bilateral amygdala and insula by lorazepam during emotion processing. Arch Gen Psychiatry 2005;62:282288. 122. Furmark T, Appel L, Michelgard A, Wahlstedt K, Ahs F, Zancan S, Jacobsson E, Flyckt K, Grohp M, Bergstrom M, Pich EM, Nilsson LG, Bani M, Langstrom B, Fredrikson M. Cerebral blood ow changes after treatment of social phobia with the neurokinin-1 antagonist GR205171, citalopram, or placebo. Biol Psychiatry 2005;58:132142. 123. Wise RG, Rogers R, Painter D, Bantick S, Ploghaus A, Williams P, Rapeport G, Tracey I. Combining fMRI with a pharmacokinetic model to determine which brain areas activated by painful stimulation are specically modulated by remifentanil. Neuroimage 2002;16:999 1014. 124. Iannetti GD, Zambreanu L, Wise RG, Buchanan TJ, Huggins JP, Smart TS, Vennart W, Tracey I. Pharmacological modulation of pain-related brain activity during normal and central sensitization states in humans. Proc Natl Acad Sci U S A 2005;102: 1819518200. 125. deCharms RC, Maeda F, Glover GH, Ludlow D, Pauly JM, Soneji D, Gabrieli JD, Mackey SC. Control over brain activation and pain learned by using real-time functional MRI. Proc Natl Acad Sci U S A 2005;102:18626 18631. 126. Morgan V, Pickens D, Gautam S, Kessler R, Mertz H. Amitriptyline reduces rectal pain related activation of the anterior cingulate cortex in patients with irritable bowel syndrome. Gut 2005; 54:601 607. 127. Berman SM, Chang L, Suyenobu B, Derbyshire SW, Stains J, FitzGerald L, Mandelkern M, Hamm L, Vogt B, Naliboff BD, Mayer EA. Condition-specic deactivation of brain regions by 5-HT3 receptor antagonist alosetron. Gastroenterology 2002; 123:969 977. 128. Drossman DA, Toner BB, Whitehead WE, Diamant NE, Dalton CB, Duncan S, Emmott S, Proftt V, Akman D, Frusciante K, Le T, Meyer K, Bradshaw B, Mikula K, Morris CB, Blackman CJ, Hu Y, Jia H, Li JZ, Koch GG, Bangdiwala SI. Cognitive-behavioral therapy versus education and desipramine versus placebo for moderate to severe functional bowel disorders. Gastroenterology 2003;125:19 31. 129. Lackner JM, Mesmer C, Morley S, Dowzer C, Hamilton S. Psychological treatments for irritable bowel syndrome: a system-

1942

MAYER ET AL

GASTROENTEROLOGY Vol. 131, No. 6

130.

131.

132.

133.

134.

135. 136.

137.

atic review and meta-analysis. J Consult Clin Psychol 2004;72:1100 1113. Pattinson KT, Rogers R, Mayhew SD, Tracey I, Wise RG. Pharmacological fMRI: Measuring opioid effects on the BOLD response to hypercapnia. J Cereb Blood Flow Metab 2006;May 31 [Epub ahead of print]. Kilpatrick L, Labus J, Berman SM, Suyenobu B, Stains J, Naliboff BD, Mayer EA. A course of tegaserod treatment modulates CNS processing of visceral afferent information. Gastroenterology 2006;130:A289 A290. Drossman DA, Ringel Y, Vogt BA, Leserman J, Lin W, Smith JK, Whitehead W. Alterations of brain activity associated with resolution of emotional distress and pain in a case of severe irritable bowel syndrome. Gastroenterology 2003;124:754 761. Naliboff BD, Derbyshire SWG, Munakata J, Berman S, Mandelkern M, Chang L, Mayer EA. Evidence for decreased activation of central fear circuits by expected aversive visceral stimuli in IBS patients. Gastroenterology 2000;118:A137. Lackner JM, Lou Coad M, Mertz HR, Wack DS, Katz LA, Krasner SS, Firth R, Mahl TC, Lockwood AH. Cognitive therapy for irritable bowel syndrome is associated with reduced limbic activity, GI symptoms, and anxiety. Behav Res Ther 2006;44:621 638. [Epub 2005; July 21]. Wise RG, Tracey I. The role of fMRI in drug discovery. J Magn Reson Imaging 2006;23:862 876. Meyer-Lindenberg A, Weinberger DR. Intermediate phenotypes and genetic mechanisms of psychiatric disorders. Nat Rev Neurosci 2006;7:818 827. Bergstrom M, Hargreaves RJ, Burns HD, Goldberg MR, Sciberras D, Reines SA, Petty KJ, Ogren M, Antoni G, Langstrom B, Eskola O, Scheinin M, Solin O, Majumdar AK, Constanzer ML, Battisti WP, Bradstreet TE, Gargano C, Hietala J. Human positron emission tomography studies of brain neurokinin 1

138. 139.

140.

141.

142.

143.

receptor occupancy by aprepitant. Biol Psychiatry 2004;55: 10071012. Craig AD. Pain mechanisms: Labeled lines versus convergence in central processing. Annu Rev Neurosci 2003;26:130. Matthews PM, Honey GD, Bullmore ET. Applications of fMRI in translational medicine and clinical practice. Nat Rev Neurosci 2006;7:732744. Hariri AR, Drabant EM, Munoz KE, Kolachana BS, Mattay VS, Egan MF, Weinberger DR. A susceptibility gene for affective disorders and the response of the human amygdala. Arch Gen Psychiatry 2005;62:146 152. Pezawas L, Meyer-Lindenberg A, Drabant EM, Verchinski BA, Munoz KE, Kolachana BS, Egan MF, Mattay VS, Hariri AR, Weinberger DR. 5-HTTLPR polymorphism impacts human cingulate-amygdala interactions: A genetic susceptibility mechanism for depression. Nat Neurosci 2005;8:828 834. Wexler BE, Gottschalk CH, Fulbright RK, Prohovnik I, Lacadie CM, Rounsaville BJ, Gore JC. Functional magnetic resonance imaging of cocaine craving. Am J Psychiatry 2001;158:86 95. Koeppe C, Schneider C, Thieme K, Mense S, Stratz T, Muller W, Flor H. The inuence of the 5-HT3 receptor antagonist tropisetron on pain in bromyalgia: a functional magnetic resonance imaging pilot study. Scand J Rheumatol Suppl 2004;119:24 27.

Received September 26, 2006. Accepted October 18, 2006. Address requests for reprints to: Emeran A. Mayer, MD, Neuroimaging Program, UCLA, Center for Neurovisceral Sciences & Womens Health, VAGLAHS, Building 115, Room 223, 11301 Wilshire Boulevard, Los Angeles, California 90073. e-mail: emayer@ucla.edu; fax: (310) 794-2864. Supported by NIH grants P50 DK64539 (to E.A.M.), R24 AT002681 (to E.A.M.), R01 DK48351 (to E.A.M.), R01 NR007768 (to B.D.N.), and R01 NS40413 (to A.D.C.).