The polyvagal theory: phylogenetic substrates of a social ... · The polyvagal theory: phylogenetic substrates of a social nervous system Stephen W. Porges ... dle-ear muscles e.g

Ž .International Journal of Psychophysiology 42 2001 123�146

The polyvagal theory: phylogenetic substrates of asocial nervous system

Stephen W. Porges

Department of Psychiatry, Uni�ersity of Illinois at Chicago, 1601 W. Taylor Street, Chicago, IL 60612-7327, USA

Received 16 October 2000; received in revised form 15 January 2001; accepted 22 January 2001

Abstract

The evolution of the autonomic nervous system provides an organizing principle to interpret the adaptivesignificance of physiological responses in promoting social behavior. According to the polyvagal theory, thewell-documented phylogenetic shift in neural regulation of the autonomic nervous system passes through threeglobal stages, each with an associated behavioral strategy. The first stage is characterized by a primitive unmyelinatedvisceral vagus that fosters digestion and responds to threat by depressing metabolic activity. Behaviorally, the firststage is associated with immobilization behaviors. The second stage is characterized by the sympathetic nervoussystem that is capable of increasing metabolic output and inhibiting the visceral vagus to foster mobilizationbehaviors necessary for ‘fight or flight’. The third stage, unique to mammals, is characterized by a myelinated vagusthat can rapidly regulate cardiac output to foster engagement and disengagement with the environment. Themammalian vagus is neuroanatomically linked to the cranial nerves that regulate social engagement via facialexpression and vocalization. As the autonomic nervous system changed through the process of evolution, so did theinterplay between the autonomic nervous system and the other physiological systems that respond to stress, includingthe cortex, the hypothalamic�pituitary�adrenal axis, the neuropeptides of oxytocin and vasopressin, and the immunesystem. From this phylogenetic orientation, the polyvagal theory proposes a biological basis for social behavior andan intervention strategy to enhance positive social behavior. � 2001 Elsevier Science B.V. All rights reserved.

Keywords: Vagus; Respiratory sinus arrhythmia; Evolution; Autonomic nervous system; Cortisol; Oxytocin; Vasopressin; Polyvagaltheory; Social behavior

1. Introduction

Embedded in the mammalian nervous systemare neuroanatomical structures related to the ex-

pression and experience of social and emotionalbehavior. Several of these structures are sharedwith other vertebrates and represent the productof phylogenetic development. Comparative re-

0167-8760�01�$ - see front matter � 2001 Elsevier Science B.V. All rights reserved.Ž .PII: S 0 1 6 7 - 8 7 6 0 0 1 0 0 1 6 2 - 3

( )S.W. Porges � International Journal of Psychophysiology 42 2001 123�146124

search across vertebrate classes provides evidencethat mammalian stress and coping responsestrategies are hierarchically ordered according to

Ž .a phylogenetic stage Porges, 1997, 1998 . Al-though research has demonstrated the relevanceof neuroanatomical structures and neurophysio-logical processes to the expression of emotionand the regulation of social behavior, contem-porary theories have not incorporated an ‘evolu-tion’ perspective.

2. Evolution as an organizing principle

Evolutionary forces have molded both contem-porary physiology and behavior. The mammaliannervous system is a product of evolution. Viaevolutionary processes, the mammalian nervoussystem has emerged with specific features thatreact to challenge in order to maintain visceralhomeostasis. These reactions change physiologi-cal state and, in mammals, limit sensory aware-ness, motor behaviors, and cognitive potentials.We can intuitively grasp the limitations of behav-ior when described in terms such as dexterity,speed and strength. However, we have difficultieswhen the limitations are manifested in the neuralregulation of physiological state. The regulationof visceral state is an important substrate of be-havior that has been both a product of and con-tributor to our evolutionary adaptive success. Thephylogeny of vertebrates illustrates a progressiveincrease in the complexity of the neural mecha-nisms available to regulate neurobehavioral stateto deal with the challenges along a continuum,defined by survival on one end, and positive so-cial-emotional experiences on the other.

To survive, mammals must determine friendfrom foe, evaluate whether the environment issafe, and communicate with their social unit.These survival-related behaviors are associatedwith specific neurobehavioral states that limit theextent to which a mammal can be physically ap-proached and whether the mammal can commu-nicate or establish new coalitions. Thus, environ-mental context can influence neurobehavioralstate, and neurobehavioral state can limit a mam-mal’s ability to deal with the environmental chal-

lenge. This knowledge of how the mammaliannervous system changes neurobehavioral states toadapt to challenge provides us with an opportu-nity to design living environments, which mayfoster both the expression of positive social be-havior and physical health.

3. The social engagement system

Through stages of phylogeny mammals and es-pecially primates have evolved a functional neuralorganization that regulates visceral state to sup-port social behavior. A social engagement system,which focuses only on the neural regulation ofthe striated muscles of the face and head and thespecific autonomic functions mediated by themyelinated vagus, has been proposed as part of amore global social nervous system. Although otherneural systems and behaviors are involved in so-cial behavior, such as signaling with trunk andlimbs, the proposed social engagement system isconceptualized to emphasize a system that has acommon neural substrate composed of severalcranial nerves that develop embryologically

Ž .together see Porges, 1998 .The social engagement system has a control

Ž .component in the cortex upper motor neuronsŽthat regulates brainstem nuclei i.e. lower motor

neurons controlling special visceral efferent path-. Ž .ways to control: eyelid opening e.g. looking ;

Ž .facial muscles e.g. emotional expression ; mid-Ždle-ear muscles e.g. extracting human voice from. Žbackground noise ; muscles of mastication e.g.

. Žingestion ; laryngeal and pharyngeal muscles e.g..vocalization and language ; and head-turning

Ž .muscles e.g. social gesture and orientation . Col-lectively, these muscles both regulate social en-gagement and modulate the sensory features ofthe environment. The neural control of thesemuscles contributes to the richness of both socialexpressions and social experiences. Important toa psychophysiological perspective, the source nu-clei of these nerves located in the brainstemcommunicate directly with the visceromotor por-tion of the nucleus ambiguus. Functionally, thevisceromotor portion of the nucleus ambiguusprovides the source nuclei for an inhibitory com-

( )S.W. Porges � International Journal of Psychophysiology 42 2001 123�146 125

ponent of the autonomic nervous system, whichcommunicates via the myelinated vagal efferentsto target peripheral organs, including the sino-atrial node. This inhibitory system promotes calmstates consistent with the metabolic demands ofgrowth and restoration by slowing heart rate,lowering blood pressure, and inhibiting sympa-thetic activation at the level of the heart.

The social engagement system is intimately re-lated to stress reactivity. In addition, the anatomi-cal structures involved in the social engagementsystem have neurophysiological interactions with

Ž .the hypothalamic�pituitary�adrenal HPA axis,the neuropeptides of oxytocin and vasopressin,and the immune system. Thus, the social nervoussystem provides a theoretical model to explain theinteractive and stress-related functions of severalphysiological systems that have central regulatorycomponents, but are expressed in the periphery.

The social nervous system functions from birthand rapidly develops to support communicationwith the environment. For example, when ahealthy infant encounters the caregiver’s face, theinfant will attempt to engage via facial expressionand vocalizations. The infant may use vocaliza-

Ž . Ž .tions cry and facial expressions grimace tosignal negative states to the caregiver. Or, tosignal more positive states, a wide-eyed, smilinginfant would attempt to elicit positive vocaliza-tions and smiles from the caregiver. Even in theyoung infant, the social engagement system ex-pects face-to-face interactions, with contingent

Žfacial expressions and vocalizations. Studies e.g..Bazhenova et al., in press have demonstrated

that when the face of the caregiver is not respon-sive, the infant will initially attempt to socially

Žengage the caregiver with display behaviors e.g..vocalizations, facial expressions . If the infant is

unsuccessful in engaging the caregiver, the infantwill become agitated and may, in the case ofhaving a depressed mother, exhibit symptoms ofdepression.

4. Neurophysiology of stress: limitations ofarousal theory

For over a century, researchers have measured

Žautonomic variables e.g. heart rate, palmar.sweat-gland activity as indicators of emotional

Žstate related to perceived stress e.g. fear, mental.effort, workload, anxiety . Interest in measuring

heart rate and sweat gland activity was theoreti-cally supported by acceptance of the arousal the-ory. Arousal theory made the assumption thatperipheral physiological measures regulated bythe sympathetic branch of the autonomic nervoussystem provided sensitive indicators of brainarousal or activation. This view was based on arudimentary understanding of the autonomic ner-vous system, in which changes in easily measured

Ž .peripheral organs e.g. sweat glands, heart wereassumed to be accurate indicators of how thebrain is processing emotional stimuli. Usually, theemotional states were associated with fight�flightbehaviors and the sympathetic-adrenal systemŽe.g. increases in heart rate, sweat gland activity,

.and circulating catecholamines as described byŽ .Cannon 1928 . The current emphasis on cortisol

as a dependent variable in stress research is con-sistent with historical views of stress-relatedarousal involving the adrenal gland.

Such measures, due in part to their availabilityand their neuroanatomical association with thetarget organs of the sympathetic nervous systemŽ .sweat glands, heart rate and with secretions

Žfrom the adrenal medulla epinephrine and nore-. Ž .pinephrine and the adrenal cortex cortisol , have

become the primary variables used to assess phys-iological reactivity. Not by plan, but by default, anarousal-theory emphasis created a research envi-ronment that minimized the importance of neuro-

Ž .physiological processes e.g. feedback and neu-Ž .roanatomical structures e.g. brain in the regula-

tion of autonomic function. Thus, there was littleinterest in several important research questions,including evolution of the neuroregulation of au-tonomic function and how the autonomic nervoussystem interacted with the immune system, theHPA axis, and the neuropeptides, oxytocin andvasopressin.

An emphasis on a global construct of ‘arousal’still abides within various sub-disciplines of psy-chology, psychiatry and physiology. This outdatedview of ‘arousal’ may restrict an understanding ofhow the autonomic nervous system interfaces with


the environment and the contribution of the au-tonomic nervous system to psychological and be-havioral processes. In contrast, more recent neu-rophysiological data promote a more integrativeview of the autonomic nervous system.

The flexibility and variability of autonomic ner-vous system function is totally dependent uponthe structure of the nervous system. By mappingthe phylogenetic development of the structuresregulating autonomic function, it is possible toobserve the dependence of autonomic reactivityon the evolution of the underlying structure ofthe nervous system, and consequently the depen-dence of social behavior on autonomic reactivity.The phylogenetic approach highlights a shift inbrainstem and cranial nerve morphology, andfunctions from a digestive and cardiopulmonarysystem to a system that integrates the regulationof facial muscles, cardiac output and the vocalapparatus for affective communication.

5. The evolution of the autonomic nervous system:emergent structures and the expression ofemotions

Although there is an acceptance that the au-tonomic nervous system and the face play a rolein emotional expression and social behavior, thereis great uncertainty regarding the autonomic sig-nature of specific or discrete emotions and thefunction of the autonomic nervous system in reg-ulating social behavior. Most researchers evaluat-ing autonomic responses during affective experi-ences, assumed, as did Cannon, that the sympa-thetic nervous system was the determinant ofemotion, or at least the primary physiologicalcovariate of emotion and an index of stress. This,of course, neglects the potential role of theparasympathetic nervous system and its neuro-physiological affinity to structures of the face andhead that regulate facial expression, eye move-ments, pupil dilation, salivation, swallowing, vo-calizing, hearing and breathing. By investigatingthe evolution of the autonomic nervous system,we may gain insight into the interface betweenautonomic function and facial expression. In thefollowing sections, the phylogenetic development

of the autonomic nervous system will be used asan organizing principle to categorize affective ex-periences.

The polyvagal theory is derived from investiga-tions of the evolution of the autonomic nervoussystem. The theory includes assumptions that im-pact on psychological, behavioral and physiologi-cal processes associated with emotional regula-tion and social behavior. Thus, consistent with thepolyvagal theory, behavioral and autonomic re-sponses associated with emotional regulation andsocial behavior reflect adaptive strategies emer-gent from the phylogeny of the mammalian ner-vous system.

1. Evolution has modified the structures of theautonomic nervous system.

2. The mammalian autonomic nervous systemretains vestiges of phylogenetically older au-tonomic nervous systems.

3. Emotional regulation and social behavior arefunctional derivatives of structural changes inthe autonomic nervous system due to evolu-tionary processes.

4. In mammals, the autonomic nervous systemresponse strategy to challenge follows a phy-logenetic hierarchy, starting with the neweststructures and, when all else fails, reverting tothe most primitive structural system.

5. The phylogenetic stage of the autonomic ner-vous system determines affective states andthe range of social behavior.

6. Polyvagal theory: three phylogenetic systems

Ž .The polyvagal theory Porges, 1995, 1997, 1998emphasizes the phylogenetic origins of brainstructures that regulate social and adaptive sur-vival-oriented defensive behaviors. The polyvagaltheory proposes that the evolution of the mam-malian autonomic nervous system provides theneurophysiological substrates for the emotionalexperiences and affective processes that are ma-jor components of social behavior. The theoryproposes that physiological state limits the rangeof behavior and psychological experience. In this


context, the evolution of the nervous system de-termines the range of emotional expression, qual-ity of communication, and the ability to regulatebodily and behavioral state. The polyvagal theorylinks the evolution of the autonomic nervous sys-tem to affective experience, emotional expression,facial gestures, vocal communication and contin-gent social behavior. Thus, the theory provides aplausible explanation of social, emotional andcommunication behaviors and disorders. The the-ory also provides an explanation of stress-relatedresponses.

The polyvagal construct was introduced to em-phasize and document the neurophysiological andneuroanatomical distinction between two

Ž .branches of the tenth cranial nerve vagus and topropose that each vagal branch was associatedwith a different adaptive behavioral strategy. Thevagus nerve, a primary component of the au-tonomic nervous system, exits the brainstem andhas branches that regulate the striated muscles of

Žthe head and face e.g. facial muscles, eyelids,middle-ear muscles, larynx, pharynx, muscles of

. Žmastication and in several visceral organs e.g..heart, gut . The theory proposes that the different

branches are related to unique, adaptive behav-ioral strategies and articulates three phylogeneticstages of the development of the mammalian

Ž .autonomic nervous system see Table 1 . Thesestages reflect the emergence of three distinctsubsystems, which are phylogenetically ordered

Žand behaviorally linked to communication e.g..facial expression, vocalization, listening ,

Ž .mobilization e.g. fight�flight behaviors and im-Žmobilization e.g. feigning death, behavioral

.‘shutdown’ and syncope . The mobilization system

is dependent on the functioning of the sympa-thetic nervous system. The most phylogeneticallyprimitive component, the immobilization system,is dependent on the unmyelinated or ‘vegetative’vagus, which is shared with most vertebrates. Withincreased neural complexity due to phylogeneticdevelopment, the organism’s behavioral and af-fective repertoire is enriched. The current paperexpands the polyvagal theory to include responsesystems that interact with the autonomic nervous

Žsystem i.e. neuropeptides, HPA axis and the im-.mune system to mediate physiological state in

order to promote or limit social behavior.Table 2 illustrates the phylogenetic differences

in the structures that regulate the heart in verte-Žbrates Taylor, 1992; Morris and Nilsson, 1994;

.Santer, 1994 . The heart is selected, because regu-lation of the heart determines the availability ofthe metabolic resources required for mobiliza-tion, as well as for growth and restoration. Forexample, cardiac output must be regulated toremain calm in safe environments, to mobilize forfight or flight behaviors, or to immobilize forfeigning death or avoidance behaviors. To regu-late cardiac output, several efferent structureshave evolved. These structures represent twoglobal, and often opposing, systems: one, a sym-pathetic-catecholamine system, including cate-cholamine-secreting chromaffin tissue and spinalsympathetic nerves; and secondly, a vagal systemŽa component of the parasympathetic nervous sys-

.tem with branches originating in medullaryŽsource nuclei i.e. dorsal motor nucleus of the

.vagus and nucleus ambiguus .In the jawless fish, neural control of the heart

is very primitive. Some jawless fish, such as hag-

Table 1The three phylogenetic stages of the neural control of the heart proposed by the polyvagal theory

Phylogenetic ANS component Behavioral function Lower motor neuronsstage

III Myelinated vagus Social communication, self- Nucleus ambiguussoothing and calming, inhibitsympathetic-adrenal influences

Ž .II Sympathetic-adrenal Mobilization active avoidance Spinal cordŽI Unmyelinated vagus Immobilization feigning death, Dorsal motor nucleus

.passive avoidance of the vagus


Table 2Neural regulation of the heart as a function of vertebratephylogeny

Group CHM DVC SNS AD�m VVC

Ž .Jawless fish X� X�Cartilaginous fish X� X�Bony fish X� X� X�Amphibians X� X� X�Reptiles X� X� X� X�Mammals X� X� X� X� X�

Abbre�iations: CHM, chromaffin tissue; DVC, dorsal vagalcomplex with vagal efferent pathways originating in the dorsalmotor nucleus of the vagus and vagal afferents terminating in

Ž .the nucleus of the solitary tract NTS ; SNS, spinal sympa-thetic nervous system; AD�m, adrenal medulla; and VVC,ventral vagal complex with efferent pathways originating in

Žthe nucleus ambiguus that regulate visceral structures heart,.bronchi, thymus and striated muscles via special visceral

efferents and afferents via the solitary tract, trigeminal andŽfacial nerve. X� indicates a cardioexcitatory influence e.g.

.increases in heart rate . X� indicates a cardioinhibitory in-Ž .fluence e.g. decreases in heart rate .

fish, rely on circulating catecholamines from dif-fuse chromaffin tissue to provide excitatory in-fluences on the heart. Other jawless fish, such aslampreys, have a cardiac vagus. However, in con-trast to all other vertebrates, which have acardio-inhibitory vagus that acts via muscariniccholinoceptors, the lamprey vagus is cardio-exci-tatory and acts via nicotinic receptors.

The cartilaginous fish, such as the sharks, raysand skates, have an unmyelinated cardio-inhibi-tory vagus. Similar to the more recent verte-brates, the vagus has cells of origin in the dorsalmotor nucleus of the vagus located in the medulla.The vagus in these fish is inhibitory and the

Žcholinoceptors the primary neurotransmitter of.the vagus in all vertebrates is acetylcholine on

the heart are muscarinic, as they are in othervertebrates. The cardioinhibitory vagus is functio-nal in the cartilaginous fish. In conditions ofhypoxia, the metabolic output is adjusted by low-ering heart rate. This modification of neural regu-lation may provide a mechanism to enable thecartilaginous fish to increase their territorialrange, by providing a neural mechanism that ad-justs metabolic output to deal with changes in

water temperature and oxygen availability. How-ever, unlike the phylogenetically more recent bonyfish and tetrapods, the cartilaginous fish do nothave direct sympathetic input to the heart. In-stead, cardiac acceleration and increases in con-tractility are mediated via �-adrenergic receptorsstimulated by circulating catecholamines releasedfrom chromaffin tissue. Thus, since activation ofmetabolic output is driven by circulating cate-cholamines, and not by direct neural innervation,once the excitatory system is triggered, the abilityto self-soothe or calm is limited.

The bony fish are phylogenetically the firstgroup of vertebrates in which the heart is regu-lated by both sympathetic and parasympatheticneural pathways. With opposing neural mecha-nisms from sympathetic and vagal pathways, rapidtransitory changes in metabolic output are possi-ble to support immediate changes in behaviorfrom mobilization to immobilization. In bony fish,this is observed as darting and freezing, withdirect neural components from the spinal cord viathe sympathetic chain producing increases in heartrate and contractility, and direct neural pathwaysfrom the brainstem via the vagus producing car-dio-inhibitory actions.

Amphibians, similar to the bony fish, have dualinnervation of the heart via systems, with directneural components from the spinal cord via thesympathetic chain producing increases in heartrate and contractility, and direct neural pathwaysfrom the brainstem via the vagus producing car-dioinhibitory actions.

A distinct adrenal medulla formed of chro-Ž .maffin tissue is present in reptiles Santer, 1994 .

The medulla is of ectodermal origin, from whichthe epidermis, nervous tissue and, in vertebrates,sense organs develop. Neural regulation by sym-pathetic nerves of the adrenal medulla provides amechanism for rapid and controlled release ofepinephrine and norepinephrine to increase car-diac output to match the metabolic demands ofmobilization behaviors. In bony fish, chromaffintissue is primarily found in parts of the cardiovas-cular system, although there is also chromaffintissue associated with the kidney. In amphibians,chromaffin tissue is primarily associated with the


kidney, and there are substantial aggregations ofchromaffin cells located along the sympatheticchain ganglia.

In mammals, the morphology of the vaguschanges. Unlike other vertebrates with a cardioin-hibitory vagus, the mammalian vagus contains twobranches. One branch originates in the dorsalmotor nucleus of the vagus and provides theprimary neural regulation of subdiaphragmaticorgans, such as the digestive tract. However, atthe level of the heart, the dorsal motor nucleus ofthe vagus does not play a major role in thenormal dynamic regulation of cardiac output.Rather, during embryological development inmammals, cells from the dorsal motor nucleus ofvagus migrate ventrally and laterally to the nu-

Ž .cleus ambiguus Schwaber, 1986 . There they formthe cell bodies for visceromotor myelinated axonsthat provide potent inhibition of the sinoatrialnode, the pacemaker for the heart. Three phylo-genetic principles can be extracted from Table 2.First, there is a phylogenetic shift in the regula-tion of the heart from endocrine communicationto unmyelinated nerves, and finally to myelinatednerves. Second, there is a development of oppos-ing neural mechanisms of excitation and inhibi-tion to provide rapid regulation of gradedmetabolic output. Third, with increased corticaldevelopment, the cortex exhibits greater control

Ž .over the brainstem via direct e.g. corticobulbarŽ .and indirect e.g. corticoreticular neural path-

ways, originating in motor cortex and terminatingin the source nuclei of the myelinated motor

Žnerves emerging from the brainstem e.g. specificvisceral neural pathways embedded within cranial

.nerves V, VII, IX, X and XI .These phylogenetic principles provide a basis

for speculations regarding the behavioral andphysiological responses associated with mam-malian social and emotional behavior, which isneurophysiologically and behaviorally linked toadaptive stress and coping strategies. In general,phylogenetic development results in increasedneural control of the heart via the myelinatedmammalian vagal system, which can promotetransitory mobilization and the expression of sym-pathetic tone without requiring sympathetic or

adrenal activation. With this new vagal system,transitory incursions into the environment orwithdrawals from a potential predator can beinitiated without the severe biological cost of themetabolic excitation associated with sympathetic-adrenal activation. Paralleling this change in neu-ral control of the heart is an enhanced neuralcontrol of the face, larynx, and pharynx that en-ables complex facial gestures and vocalizations.This phylogenetic course results in greater centralnervous system regulation of behavior, especiallybehaviors needed to engage and disengage withenvironmental challenges. Birds are conspicu-ously missing from Table 2, since Table 2 includesthe phylogenetic stage preceding mammals. Al-though mammals and birds are phylogenetic de-scendants of reptiles, mammals are not ‘direct’phylogenetic descendants of birds.

By transitory down-regulation of the cardioin-Žhibitory vagal tone to the heart i.e. removing the

.�agal brake , the mammal is capable of rapidincreases in cardiac output without activating thesympathetic-adrenal system. By engaging this sys-tem, rather than the sympathetic-adrenal system,mammals have the opportunity to rapidly increasemetabolic output for immediate mobilization. Un-der prolonged challenge, the sympathetic systemmay also be activated. However, by rapidly re-en-gaging the vagal system, mammals have the ca-pacity to inhibit sympathetic input to the heartŽ .Vanhoutte and Levy, 1979 and rapidly decreasemetabolic output to self-soothe and calm.

7. The vagal brake

The myelinated mammalian vagus is activelyinhibitory of the sympathetic nervous system atthe level of the heart. The mammalian vagus may

Žfunction as an active �agal brake see Porges et.al., 1996 in which rapid inhibition and disinhibi-

tion of the vagal tone to the heart can rapidlymobilize or calm an individual. In addition, sincethe mammalian vagus has distinct pathways in-volved in the �oluntary regulation of the striated

Žmuscles e.g. corticobulbar pathways, afferentsfrom face and mouth, efferents to larynx and


.pharynx , the regulation of heart rate is neu-roanatomically and neurophysiologically linked tothe function of the special visceral efferents.

Due to the tonic vagal influences to the sinoa-Ž .trial node the heart’s pacemaker , resting heart

rate is substantially lower than the intrinsic rateof the pacemaker. When the vagal tone to thepacemaker is high, the vagus acts as a brake onthe rate at which the heart is beating. When vagaltone to the pacemaker is low, there is little or noinhibition of the pacemaker. Thus, the vagal brakemay be used as a construct to describe the func-tional modulation of heart rate by the myelinatedvagal efferent pathways. The vagal brake providesa neural mechanism to rapidly change visceralstate by slowing or speeding the heart rate. Con-sistent with the assumptions of the polyvagal the-ory, the vagal brake contributes to the modula-tion of cardiac output by decreasing the inhibitoryvagal control of the heart to speed heart rate andby increasing the inhibitory vagal control of theheart to slow heart rate. Thus, neurophysiologi-cally, the vagal brake provides a mechanism tosupport the metabolic requirements for mobiliza-tion and communication behaviors. Functionally,the vagal brake , by modulating visceral state,enables the individual to rapidly engage and dis-engage objects and other individuals and to pro-mote self-soothing behaviors and calm behavioralstates.

Difficulties in regulating the vagal brake resultin the recruitment of other neural mechanismsŽ .e.g. sympathetic regulation of the heart and

Žneural chemical mechanisms e.g. stimulation of.the HPA axis to regulate physiological state.

Thus, consistent with the polyvagal theory, if thevagal brake is not functioning or will not servethe survival needs of the organism, the phyloge-

Žnetically older systems e.g. the sympathetic-.adrenal system will be recruited to regulate

metabolic output to deal with environmental chal-lenges. For example, if the vagal brake is notfunctioning, there is the potential for greaterdependence on the sympathetic excitation of thecardiovascular system with the associated health-

Ž . Žrelated e.g. hypertension and behavioral e.g..irritable and reactive costs.

8. Evolution and dissolution: a hierarchicalresponse strategy

The evolution of the autonomic nervous systemprovides substrates for the emergence of threeadaptive stress and coping subsystems, each linkedto structures that evolved during identifiable phy-logenetic stages. The polyvagal theory proposesthat during danger or threat the older, less socialsystems are recruited. The older systems, al-though functional in the short term, may result indamage to the mammalian nervous system whenexpressed for prolonged periods. Thus, the stressand coping neurophysiological strategies that are

Žadaptive for reptiles e.g. apnea, bradycardia, im-.mobilization may be lethal for mammals.

Rather than describing the autonomic nervoussystem as a linear ‘arousal’ system focused on thesympathetic nervous system, or a balance systemfocused on the opposing influences of the sympa-thetic and parasympathetic pathways, the func-tion of the autonomic nervous system is hierarchi-cally organized. As emphasized above, the hierar-chical organization is phylogenetically determinedand can be summarized as the following threesequential functional subsystems:

Ž .1. The ventral vagal complex VVC : a mam-malian signaling system for motion, emotion,and communication.

Ž .2. The sympathetic nervous system SNS : anadaptive mobilization system supporting fightor flight behaviors.

Ž .3. The dorsal vagal complex DVC : a vestigialimmobilization system.

Each of these three neural constructs is linkedwith a specific response strategy observable inhumans. Each strategy is manifested via differen-tiated motor output from the central nervoussystem to perform specific adaptive functions: toimmobilize and conserve metabolic resourcesŽ .DVC , to mobilize in order to obtain metabolic

Ž .resources SNS , or to signal with minimal energyŽ .expense VVC . The constituent responses associ-

ated with each subsystem are listed in Table 1.


( )8.1. The �entral �agal complex VVC

The primary efferent fibers of the VVC origi-nate in the nucleus ambiguus. The primary af-ferent fibers of the VVC terminate in the sourcenuclei of the facial and trigeminal nerves. TheVVC has the primary control of supradiaphrag-matic visceral organs, including the larynx, phar-ynx, bronchi, esophagus, and heart. Motor path-

Žways from the VVC to visceromotor organs e.g..heart and bronchi and somatomotor structures

Ž .e.g. larynx, pharynx, esophagus are myelinatedto provide tight control and speed in responding.In mammals, the visceromotor fibers to the heartexpress high levels of tonic control and are capa-ble of rapid shifts in cardioinhibitory tone toprovide dynamic changes in metabolic output tomatch environmental challenges. This rapid regu-lation characterizes the qualities of the mam-malian vagal brake that enable rapid engagementand disengagement in the environment withoutmobilizing the sympathetic nervous system.

A major characteristic of the VVC is the factthat the neural fibers regulating somatomotorstructures are derived from the branchial orprimitive gill arches that evolved to form cranialnerves V, VII, IX, X and XI. Somatomotor fibersoriginating in these cranial nerves control thebranchiomeric muscles, including those of theface, mastication, neck, larynx, pharynx, esopha-gus, and middle ear. Visceromotor efferent fiberscontrol salivary and lachrymal glands, as well asthe heart and bronchi. The primary afferents tothe VVC come from facial and oral afferentstraveling through the facial and trigeminal nervesand the visceral afferents terminating in the nu-

Ž .cleus of the solitary tract NTS . The VVC isinvolved in the control and coordination of suck-ing, swallowing, and vocalizing with breathing.

( )8.2. The sympathetic ner�ous system SNS

The sympathetic nervous system is primarily asystem of mobilization. It prepares the body foremergency by increasing cardiac output, stimulat-ing sweat glands to protect and lubricate the skin,and by inhibiting the metabolically costly gas-trointestinal tract. The evolution of the sympa-

thetic nervous system follows the segmentation ofthe spinal cord, with cell bodies of the pregan-glionic sympathetic motor neurons located in thelateral horn of the spinal cord. The sympatheticnervous system has long been associated withemotion and stress. The label sympathetic reflectsthe historical identity of this system as a nervoussystem with feelings and contrasts it with theparasympathetic nervous system, a label that re-flects a nervous system that guards against feel-ings.

( )8.3. The dorsal �agal complex DVC

Ž .The dorsal vagal complex DVC is primarilyassociated with digestive, taste, and hypoxic re-sponses in mammals. It includes the nucleus of

Ž .the solitary tract NTS and the interneuronalcommunication between the NTS and dorsal mo-

Ž .tor nucleus of the vagus DMX . The efferents forthe DVC originate in the DMX and primary vagalafferents terminate in the NTS. The DVC pro-vides the primary neural control of subdiaphrag-matic visceral organs. It provides low tonic influ-ences on the heart and bronchi. This low tonicinfluence is the vestige from the reptilian vagalcontrol of the heart and lung. In contrast toreptiles, mammals have a great demand for oxy-gen and are vulnerable to any depletion in oxygenresources.

The DVC provides inhibitory input to thesinoatrial node of the heart via unmyelinatedfibers, and thus is less tightly controlled than themyelinated fibers from the VVC. Hypoxia or per-ceived losses of oxygen resources appear to be themain stimuli that trigger the DVC. This responsestrategy is observed in the hypoxic human fetus.Although adaptive for the reptile, the hypoxictriggering of this system may be lethal for mam-mals. In addition, it is important to note that theDVC has beneficial functions in humans. Undermost normal conditions, the DVC maintains toneto the gut and promotes digestive processes.However, if up-regulated, the DVC contributes topathophysiological conditions including the for-mation of ulcers via excess gastric secretion andcolitis. Recent research supports the importanceof the unmyelinated vagal fibers in bradycardia


Ž .Cheng and Powley, 2000 and suggests the possi-bility that massive bradycardia may be de-termined by the unmyelinated vagal fibers associ-ated with the DVC recruiting myelinated vagalfibers to maximize the final vagal surge on the

Ž .heart Jones et al., 1995 .

8.4. Dissolution

The polyvagal theory proposes a hierarchicalresponse strategy to environmental challenges,with the most recent modifications employed firstand the most primitive last. However, the re-sponse strategy is not all-or-none, and may in-clude transitional blends between the boundariesof the three hierarchical stages. These transition-al blends may be determined by both visceral

Žfeedback and higher brain structures includingthe HPA axis and vasopressinergic and oxytocin-ergic pathways that communicate between the

.hypothalamus and the DVC . Thus, the neuro-physiological substrate of specific behavioral statesand coping strategies may incorporate activationof a sequence of response systems representingmore than one phylogenetic stage.

This phylogenetically based hierarchical re-sponse strategy is consistent with the concept ofdissolution proposed by John Hughlings JacksonŽ .1958 to explain disease of the nervous system.Jackson proposed that:

Ž .The higher nervous arrangements inhibit or controlthe lower, and thus, when the higher are suddenlyrendered functionless, the lower rise in activity.

Ž .The polyvagal theory Porges, 1995, 1997, 1998proposes dissolution, not in response to diseaseor brain trauma, but as an adaptive biobehavioral

Žresponse strategy to differential challenges i.e..threats to survival . Although reminiscent of the

Ž .triune brain proposed by MacLean 1990 , thepolyvagal theory emphasizes that even the phylo-genetically more primitive structures havechanged in structure and function. This phyloge-netic adjustment of the autonomic nervous sys-

Žtem represents an exaptation i.e. a shift in func-.tion of structures to express emotions and to

adapt to the complex environment.

Expanding the Jacksonian construct of dissolu-tion from disease and trauma, the polyvagal the-ory proposes that adaptive response strategies tosurvival challenges follow a phylogenetically de-fined hierarchy. The VVC, with its mechanisms ofsignaling and communication, provides the initialresponse to the environment. The VVC inhibits,at the level of the heart, the strong mobilizationresponses of the sympathetic nervous system.Withdrawal of VVC, consistent with Jackson’smodel, results in a disinhibition of the sympa-thetic control of the heart. Similarly, withdrawalof sympathetic tone results in a disinhibition ofthe DVC control of the gastrointestinal tract andvulnerability of the bronchi and heart. There areseveral clinical consequences to unopposed DVCcontrol including: defecation, due to a relaxationof the sphincter muscles and increased motility ofthe digestive tract; apnea, due to constriction ofthe bronchi; and bradycardia, due to stimulationof the sinoatrial node. Thus, when all else fails,the nervous system elects a metabolically conser-vative course that is adaptive for primitive verte-brates. For mammals, this strategy may be adap-tive in the short term, but lethal if maintained.Consistent with the Jacksonian principle of disso-lution, specific psychopathologies defined by af-fective dysfunction may be associated with au-tonomic correlates consistent with the three phy-logenetic levels of autonomic regulation. Thethree levels do not function in an all-or-nonefashion; rather, they exhibit gradations of controldetermined by both visceral feedback and higherbrain structures.

9. The social engagement system: an emergentproperty of the ventral vagal complex

Phylogenetically, the ventral vagal complexŽ .VVC is the most recent neurophysiological af-fect system. The VVC is composed of a somato-motor component, consisting of the special vis-ceral efferent pathways, and a visceromotor com-ponent, consisting of the myelinated vagal path-ways from the nucleus ambiguus to the sinoatrialnode of the heart and the bronchi. As illustrated


Fig. 1. The social engagement system: social communication is determined by the cortical regulation of medullary nuclei viaŽcorticobulbar pathways. The social engagement system consists of a somatomotor component i.e. special visceral efferent pathways

. Žthat regulate the muscles of the head and face and a visceromotor component i.e. the myelinated vagus that regulates the heart.and bronchi . Solid arrows indicate the somatomotor component. Dashed arrows indicate the visceromotor component.

in Fig. 1, the special visceral efferents and thevagal brake collectively constitute an emergentsocial engagement system. The somatomotor com-ponents of the VVC contribute to the regulationof behaviors involved in exploration of the social

Ž .environment e.g. looking, listening, ingesting andbehaviors involved in acknowledging social con-

Ž .tact e.g. facial and head gestures, vocalizing .More specifically, the somatomotor components

Žof the VVC are involved in head-turning via. Ž .cranial nerve XI , vocalizations IX, X , facial

Ž .expression VII, V , the filtering of low-frequencysounds via the middle-ear muscles to extract hu-

Ž .man voice from background sounds VII , andŽ .mastication V . The visceromotor components of

the VVC contribute to the rapid modulation ofŽ .vagal X control of the heart and the bronchi

Ž .X , which provides metabolic resources to en-gage and disengage in a social setting.

Three important features define the social en-gagement system. First, the efferent pathways thatregulate the social engagement system originate

Žin medullary structures i.e. nucleus of cranialnerve V, nucleus of cranial nerve VII, nucleus

.ambiguus . Second, corticobulbar pathways, whichŽoriginate in frontal cortex i.e. upper motor neu-

.rons , enable the possibility of efficient corticalŽregulation of these medullary source nuclei i.e.

.lower motor neurons . Third, on the medullarylevel, the structures that regulate the efferentregulation of social-communication behaviorsneuroanatomically communicate with structures

Žthat regulate ingestion e.g. sucking, swallowing,.salivation and cardiac output. Thus, modulation

of the vagal brake may either promote calmingŽand self-soothing states i.e. attenuate the influ-

ence of the sympathetic nervous system on the. Žheart or support mobilization i.e. potentiate the

influence of the sympathetic nervous system on.the heart .

Consistent with the polyvagal theory, difficul-ties in regulating the �agal brake may result in the

Žphylogenetically older systems i.e. neural regula-tion of the adrenal and the sympathetic nervous

.system being recruited to regulate metabolic out-put to deal with environmental challenges. Thus,the polyvagal theory would predict that duringstates of mobilization, characterized by classic


‘fight�flight’ behaviors and sympathetic excita-tion, both the �agal brake and the behavioralcomponents of the social engagement system wouldnot be easily accessible.

10. Measurement

In the psychophysiological literature, there isan assumption that respiratory sinus arrhythmiais an index of vagal influences to the heart. Thepolyvagal theory distinguishes between therhythmic beat-to-beat changes in heart rate, me-diated via the myelinated vagal pathways thatoriginate in the nucleus ambiguus, and the con-servative responses of bradycardia, mediated viathe unmyelinated vagal pathways that originate inthe dorsal motor nucleus of the vagus.

The nucleus ambiguus is a component of anetwork of structures that generate a respiratoryrhythm to foster coordination among cardiac and

Ž .respiratory processes Richter and Spyer, 1990 .In support of this model, vagal fibers originatingfrom the nucleus ambiguus and terminating in

Ž .both the bronchi Haselton et al., 1992 and theŽ .sinoatrial node Spyer and Jordan, 1987 have a

respiratory rhythm. In contrast, vagal fibers origi-nating from the dorsal motor nucleus of the vagusdo not have a respiratory rhythm.

Ž .Recent work by Jones et al. 1995 providessupport for this model. They note that the pre-ganglionic unmyelinated vagal fibers originatingin the dorsal motor nucleus of the vagus do nothave an obvious input from peripheral or centralrespiratory or cardiac related inputs, but are acti-vated by stimulation of pulmonary unmyelinatedafferent fibers. Convergent support comes from

Ž .Paton 1998 , who reported that stimulation ofunmyelinated vagal afferents produces reflexbradycardia and depresses central respiratory ac-tivity, consistent with an adaptive shutdown sys-tem as proposed by the polyvagal theory. Thus,the vagal responses mediated by the unmyeli-nated vagus, being part of an ancient conserva-tion system, are reflected in mammals as adaptiveresponses triggered by life-threatening events,

Ž .such as hypoxia see Potter and McCloskey, 1986 ,

and can be observed as clinical bradycardia in theŽ .fetus Reed et al., 1999 .

Neuroanatomical confirmation that the two va-gal pathways in mammals have different function-

Žal impact on the heart i.e. beat-to-beat variability. Žand level has been confirmed Jones et al., 1995;

.Cheng and Powley, 2000 . However, the functionsof the unmyelinated fibers are not totally under-stood. For example, the myelinated and unmyeli-nated fibers differentially respond to rates ofstimulation, suggesting different functional

Ž .properties Fan and Andresen, 1998 . Althoughstimulation of the unmyelinated fibers originatingin the dorsal motor nucleus of the vagus canproduce heart-rate slowing and an attenuation of

Ž .contractility see Cheng et al., 1999 , whether theunmyelinated fibers are responsible for clinical

Žbradycardia has been contested see Jones et al.,.1995; Hopkins et al., 1996 .

Alternative explanations might be considered.For example, it is possible that the myelinatedvagal fibers originating in the nucleus ambiguusmight contribute to clinical bradycardia. This ex-planation would suggest that the myelinated vagalpathways might, under different conditions, re-spond to two regulatory systems. During condi-tions when blood gases are within normal range,part of a rhythmic vagal system may produce acoordinated respiratory rhythm in heart rate andbronchial activity to facilitate oxygen diffusion.However, during conditions of compromise whenblood gas status threatens survival, both RSA andrespiratory drive would be suppressed. In theabsence of the rhythmic vagal system, a tonicvagal system may drop heart rate level to reducemetabolic demands. Perhaps during fetal distress,a tonic increase in the influence of the myelin-

Žated vagal fibers to the heart i.e. producing.bradycardia is paralleled by the increase in the

tonic influence of the unmyelinated vagal fibersŽ .to the gut i.e. producing meconium . Consistent

with this premise, there are reports that selectivestimulation of the myelinated vagal fibers inducebronchoconstriction and potentiate the recruit-

Ž .ment of non-myelinated fibers Lama et al., 1988 .Based on the above neurophysiological evi-

dence, the functional impact of the myelinated


vagus on the heart is easily calibrated by quantify-ing the amplitude of respiratory sinus arrhythmiaŽi.e. the rhythmic increase and decrease in heartrate observed at the frequency of spontaneous

.breathing . In addition, the period of the oscilla-Ž .tions in heart rate the period of RSA would

provide a valid index of the output frequency ofthe cardiopulmonary oscillator. In addition, sincethe brainstem nuclei that regulate the myelinatedvagal pathways to the heart are neuroanatomi-cally and neurophysiologically linked to the brain-stem source nuclei of the special visceral effer-ents that regulate facial expression, monitoring

Ždynamic changes in RSA and heart rate i.e. the.�agal brake provides an efficient and non-invasive

method of assessing the status of the social en-gagement system.

11. Voodoo or vagus death: a test of the polyvagaltheory

The polyvagal theory provides a theoreticalframework to interpret the phenomenon of Voo-

Ž .doo or fright death described by Cannon 1957Ž .and Richter 1957 . Cannon believed that ex-

treme emotional stress, regardless of the specificbehavioral manifestation, could be explained interms of degree of sympathetic-adrenal excita-tion. Voodoo death was assumed to be directlyattributable to emotional stress. Being wed to asympatho-adrenal model of emotional experienceŽ .see above , Cannon assumed that Voodoo deathwould be the consequence of the state of shockproduced by the continuous outpouring ofepinephrine via excitation of the sympathetic ner-vous system. According to the Cannon model, thevictim would be expected to breathe very rapidlyand have a rapid pulse. The heart would beat fastand gradually lead to a state of constant contrac-tion, and ultimately, to death in systole. Since hisspeculations were not empirically based, he of-fered the following challenge to test his model ofVoodoo death:

If in the future, however, any observer has opportunityto see an instance of voodoo death, it is to be hopedthat he will conduct the simpler tests before the victim’slast gasp.

Ž .Curt Richter 1957 responded to Cannon’schallenge with an animal model. Rats were pre-stressed, placed in a closed, turbulent water tank,and the latency to drowning was recorded. Mostdomestic laboratory rats lasted for several hours,while, unexpectedly, all of the wild rats died within15 min. In fact, several wild rats dived to thebottom, and without coming to the surface, died.To test Cannon’s hypothesis, that stress-inducedsudden death was sympathetic, Richter monitoredheart rate and determined whether the heart wasin systole or diastole after death. He assumed,based upon Cannon’s speculations, that tachycar-dia would precede death, and that at death, theheart would be in a state of systole, reflecting thepotent effects of sympathetic excitation on thepacemaker and the myocardium. However,Richter’s data contradicted the Cannon model.Heart rate slowed prior to death, and at death theheart was engorged with blood, reflecting a stateof diastole. Richter interpreted the data as de-monstrating that the rats died a vagus death, theresult of over-stimulation of the parasympatheticsystem, rather than of the sympathico-adrenalsystem. However, Richter provided no physiologi-cal explanation, except the speculation that thelethal vagal effect was related to a psychologicalstate of hopelessness.

The immediate and reliable death of the wildrats in Richter’s experiment may represent a moreglobal immobilization strategy. Sudden prolongedimmobility or feigned death is an adaptive re-sponse exhibited by many mammalian species.

Ž .Hofer 1970 demonstrated that several rodentspecies when threatened exhibited an immobilitythat was accompanied by very low heart rate. Forsome of the rodents, heart rate during immobilitywas less than 50% of the basal rate. During theprolonged immobility, respiration became so shal-low that it was difficult to observe, although therate greatly accelerated. Although physiologicallysimilar, Hofer distinguished between prolongedimmobility and feigned death. The onset offeigned death, unlike the behavior of ‘hopeless-ness’ described by Richter, occurred suddenly withan apparent motor collapse during active strug-gling. Similar to Richter, Hofer interpreted this


fear-induced slowing of heart rate as a vagalphenomenon. In support of this interpretation, henoted that of the four species that exhibited pro-longed immobility, 71% of the subjects had car-diac arrhythmias of vagal origin; in contrast, inthe two species that did not exhibit immobilitybehaviors, only 17% exhibited cardiac ar-rhythmias of vagal origin.

The polyvagal theory places Richter’s andHofer’s observations in perspective. Following theJacksonian principle of dissolution, the rodentswould exhibit the following sequence of response

Ž . Ž .strategies: 1 removal of VVC tone; 2 increaseŽ .in sympathetic tone; and 3 a surge in DVC tone.

It appears that the more docile domestic rats inRichter’s experiment progressed from a removalof VVC tone, to an increase in sympathetic tone,and then died via exhaustion. However, the pro-file of the wild rats was different. Being totallyunaccustomed to enclosures and handling, andalso having their vibrissae cut, a mobilizationstrategy driven by increased sympathetic tone wasnot functional. Instead, these rats reverted totheir most primitive system to conserve metabolicresources via DVC. This strategy promoted animmobilization response characterized by reducedmotor activity, apnea, and bradycardia. Unfortu-nately, this mode of responding, although adap-tive for reptiles, is lethal for mammals. Similarly,the onset of feigned death, as described by Hofer,illustrates the sudden and rapid transition froman unsuccessful strategy of struggling, requiringmassive sympathetic activation, to the metaboli-cally conservative immobilized state, mimickingdeath, associated with the DVC.

These data suggest that the vagus contributesto severe emotion states and may be related toemotional states of immobilization, such as ex-treme terror. Application of the polyvagal ap-proach enables the dissection of vagal processes

Ž .into three strategic programs: 1 when tone ofthe VVC is high, there is an ability to communi-cate via facial expressions, vocalizations, and ges-

Ž .tures; 2 when tone of the VVC is low, thesympathetic nervous system is unopposed andeasily expressed to support mobilization, such as

Ž .fight or flight behaviors; and 3 when tone fromDVC is high, there is immobilization and poten-

tially life-threatening bradycardia, apnea, and car-diac arrhythmias.

12. How does the polyvagal theory provide atheoretical platform to understand the functionalinteractions of the HPA axis, the neuropeptides ofoxytocin, and vasopressin with the autonomicnervous system?

12.1. Structures and functions of the adrenal gland

The adrenal gland exhibits phylogeneticchanges in structure and function. Above, be-cause of its complimentary function to the sympa-thetic nervous system, the adrenal medulla wasdiscussed. However, since the adrenal cortex istreated as an endocrine organ and part of theHPA axis, it is traditionally studied and discussedindependently of the adrenal medulla and theautonomic nervous system. Unlike the medulla,which is of ectodermal origin, the cortex is ofmesodermal origin. The adrenal cortex secretescortisol, which is frequently used to index stress.

All vertebrates produce corticosteroids and cat-echolamines. However, the structures that pro-duce these secretions follow a phylogenetic trendthat results in two anatomical changes. First, thechromaffin tissue and the adrenocortical cells,which are scattered through the viscera in theprimitive vertebrates, cluster together to formfunctional structures. Second, these structures es-

Ž .tablish a close anatomical relation the adrenalwith communication between the secretions ofthe two components of the adrenal, and only inmammals does the adrenal have its own vascularsupply and venous drainage.

Only mammals have adrenocortical cells clus-tered as a cortex of the adrenal. Interestingly,

Ž .several reptiles e.g. lizards and some snakes thathave a functioning adrenal medulla have theadrenocortical cells partially encapsulated by the

Žchromaffin cells. In the more primitive fish jaw-.less and cartilaginous , the chromaffin and

adrenocortical cells are entirely separate. In thebony fish, both the adrenocortical cells and chro-maffin tissue are scattered as independent groupswithin the ‘head’ kidney. In amphibians, the chro-


maffin and adrenocortical tissues are scatteredover the ventral surfaces of the kidney. In rep-tiles, the chromaffin and adrenocortical tissuesbecome consolidated into distinct adrenal glands.Reptiles provide a phylogenetic marker, in whichthe interrenals have an independent vascular sup-

Žply and venous drainage and no longer as in.amphibians and bony fish rely on the kidney and

a renal portal system for the distribution of secr-etory products.

Several investigators have explored the adapt-ive function of the close proximity of the cor-

Žtisol-secreting adrenocortical cells that subs-.equently form the adrenal cortex and the cat-

Žecholamine-secreting chromaffin tissue that sub-. Žsequently forms the adrenal medulla Norris,

.1997 . The medulla receives large quantities ofcorticosteroids through the adrenal vascular sys-tem, and these hormones activate the enzymesystem for converting norepinephrine intoepinephrine. This effect would support sympa-thetic-adrenal medulla functions associated withmobilization and increased metabolic activity.Epinephrine has a potent effect on the elevationof blood glucose, convergent with the influencesof some of the adrenal steroids. Additionally,cortisol may directly enhance mobilization byconverting lactate to glucose via gluconeogenesisin the liver. This process would contribute to themetabolic demands of mobilization by increasingthe availability of glucose, and also by reducingoxygen debt due to the accumulation of lactate.

Vagal activity has been implicated in the func-tion of the adrenal cortex. Reports suggest thatafferents originating in the subdiaphragmatic va-

Ž .gus i.e. DVC exhibit an inhibitory influence onŽthe HPA axis and reduce cortisol secretion e.g.

.Bueno et al., 1989; Miao et al., 1997 . Otherresearch has demonstrated a covariation betweenincreases in cortisol and decreases in cardiac va-

Ž .gal tone measured via RSA Gunnar et al., 1995 ,which, consistent with the removal of the vagalbrake and the stimulation of the sympathetic ner-vous system, would promote mobilization. Simi-larly, psychological stressors, which reduce car-diac vagal tone, increase cortisol plasma levelŽ .e.g. Cacioppo et al., 1995 . Thus, in several situa-tions there appears to be a coordinated response

that functions to promote metabolic activity andmobilization behaviors by withdrawal of VVC toneand increasing both SNS activity and activation ofthe HPA axis.

In general, functioning of the adrenal cortexand the secretion of cortisol appear to be inte-grated into the mobilization function of the au-tonomic nervous system by increasing sympa-thetic activation and circulating catecholamines.These effects suggest that, consistent with thephylogenetic approach described in the polyvagaltheory, cortisol secretion may be related to main-

Žtenance of mobilization i.e. the conversion of.norepinephrine into epinephrine for fight�flight

behaviors, and recovery from the lactate build-upthat may contribute to a functional oxygen debtŽ .i.e. gluconeogenesis .

In addition, the reports of dysregulation of theHPA axis, low cortisol or low cortisol reactivity insocial and affective disorders, such as schizophre-

Ž .nia Jansen et al., 2000 , posttraumatic stress dis-Ž .order Yehuda et al., 1996 and the consequences

Žof neglect and abuse in children De Bellis et al.,.1994 , may be explained within the context of the

polyvagal theory. According to the theory, whenŽ .mobilization strategies fight�flight behaviors are

ineffective in removing the individual from thestressor and modulating stress, then the nervoussystem may degrade to a phylogenetically earlierlevel of organization. Thus, low cortisol or a hy-poresponsive HPA axis may reflect a neural strat-

Žegy associated with immobilization e.g. passive.avoidance, death feigning, dissociative states that

would require a reduction in energy resources.

12.2. Oxytocin��asopressin

Phylogenetically, although all vertebrates havepeptides similar in structure to both oxytocin andvasopressin, only mammals have the specific re-ceptors for both peptides. Oxytocin and vaso-pressin are synthesized primarily in the paraven-tricular and supraoptic nuclei of the hypothala-mus and released centrally via parvocellular neu-rons and systemically via magnocellular neuronsŽ .Swanson and Sawchenko, 1977 . The central andsystemic effects of these neuropeptides are dif-ferent. Central release of oxytocin regulates the


output of the dorsal motor nucleus of the vagus,usually maintaining output within levels optimalto support homeostasis and providing a proposed

Ž .‘anti-stress’ function see Carter, 1998 . Periph-eral release of oxytocin is related to milk ejection,uterine contractions and ejaculation. Central re-lease of vasopressin appears to modulate afferentfeedback from the viscera and to shift set-points,independent of sensitivity, for vagal reflexes, such

Ž .as the baroreceptor reflex Michelini, 1994 . Theraising of the baroreceptor set-point, by increas-ing cardiac output, potentiates fight�flight behav-iors and allows sympathetic excitation of the heartto be unopposed by homeostatic vagal reflexes.Thus, central levels of oxytocin have been as-sumed to be associated with vagal processes, andcentral levels of vasopressin have been assumedto be associated with sympathetic processesŽ .Uvnas-Moberg, 1997 .

Because the peripheral influences of oxytocinand vasopressin function through feedback, pri-marily via the sensory component of the DVC,the peripheral effects of these peptides are lessclear and may be level-dependent, or differ as afunction of acute versus chronic exposure. Forexample, it is possible that peripheral vasopressinmay, by stimulating vagal afferents, trigger mas-sive vagal responses via the dorsal motor nucleusof the vagus. In support of this speculation, it isknown that in humans, peripheral vasopressin,and not oxytocin, is related to the nausea experi-

Ž .enced during motion sickness Koch et al., 1990 .In addition, systemic vasopressin may induce abaroreceptor-mediated bradycardia and a fall in

Žplasma concentration of norepinephrine Buwalda.et al., 1992; Michelini, 1994 .

Oxytocin may be part of a complex responseprofile related to the perception of the environ-

Ž .ment as safe see Fig. 2 . Consistent with thisŽ .view, Uvnas-Moberg 1997 and Carter and Alte-

Ž .mus 1997 propose that oxytocin promotes statesŽ .resistant to stress i.e. anti-stress . In contrast,

vasopressin may be part of a complex responseprofile related to the perception that the environ-

Ž .ment is challenging or unsafe see Fig. 3 . In fact,central vasopressin could potentiate mobilizationresponses via sympathetic excitation, while highlevels of systemic vasopressin may potentiate a

Fig. 2. Neural and neuropeptide regulation of the dorsal vagalŽ .complex DVC in a safe environment. The DVC includes

Ž .sensory nuclei in the nucleus of the solitary tract NTS andarea postrema, and motor nuclei in the dorsal motor nucleus

Ž .of the vagus DMX . During states when the individual per-Ž .ceives the environment as safe, oxytocin OT is released

centrally to the sensory and motor portions of the DVC andsystemically to the visceral organs. Functionally, OT fosters acalm immobilization or ‘anti-stress’ state.

Žphysiological shutdown associated with fear e.g..bradycardia via feedback to the dorsal motor

nucleus and inhibition of sympathetic outflowŽ .Ferguson and Lowes, 1994 . In addition, lesionsof vagal afferents, which functionally block thevisceral input to the sensory component of the

Ž .DVC areas sensitive to vasopressin , attenuate orabolish specific, conditioned taste aversionsŽ .Andrews and Lawes, 1992 .

Paralleling the phylogenetic shift in the cells oforigin and the myelination of the vagus, empha-sized in the polyvagal theory, is a modification ofthe hypothalamic regulation of the DVC via bothoxytocin and vasopressin. In mammals, the adventof specific receptors for oxytocin and vasopressinincreases the range of adaptive functions involv-ing the DVC. In mammals, the dorsal motornucleus of the vagus, the motor component of the


Fig. 3. Neural and neuropeptide regulation of the dorsal vagalŽ .complex DVC in an unsafe environment. During perceived

danger, when mobilization is adaptive, central vasopressiner-Ž .gic pathways AVP communicate between the hypothalamus

and both NTS and area postrema to change the set-point ofŽ .vagal reflexes e.g. baroreceptor reflex to facilitate sympa-

thetic excitation.

DVC, is sensitive to oxytocin and insensitive tovasopressin. In contrast, the sensory componentsof the DVC, nucleus of the solitary tract and areapostrema, are most sensitive to vasopressin. Areapostrema, a medullary structure, interacts withvagal afferents and communicates with the nu-cleus of the solitary tract. Area postrema is acircumventricular organ, which has a rich capil-lary plexus and lacks a blood brain barrier. Areapostrema plays a major role in vasopressin syn-

Ž .thesis and release Arima et al., 1998 . Since areapostrema lacks a blood brain barrier, it provides aportal for circulating peptides such as vasopressinto influence brain function. Research has demon-strated that activation of the neurons in areapostrema augments the NTS-mediated arterialbaroreflex via vasopressin administered, either bydirect microinjection into area postrema, or via

Ž .venous infusion Qu et al., 1997 . Although thenucleus of the solitary tract has receptors for

Ž .oxytocin Landgraf et al., 1990 , area postremamay not be directly influenced by oxytocinŽ .Carpenter, 1990 . The differential sensitivity ofspecific components of the DVC to these two

Žneuropeptides the differential effects of centraland systemic release on visceral function and a

.potential level dependency results in a widerrange of response options, including maximizingmobilization behaviors and co-opting of theprimitive vagal system associated with im-mobilization to support ‘anti-stress’ functions,such as social engagement and growth and

Ž .restoration see Porges, 1998 . In support of thismodel, it has been reported that oxytocinergicprojections to the DVC restrain exercise-induced

Ž .tachycardia Braga et al., 2000 .Although the polyvagal theory has emphasized

the potentially lethal shutdown behaviors associ-ated with massive surges from the dorsal motornucleus of the vagus, the DVC is involved inother functions. The DVC, with motor fibers orig-inating in the dorsal motor nucleus of the vagusand afferent fibers terminating in the nucleus ofthe solitary tract and area postrema, has beenassumed to be involved primarily in homeostatic

Ž .functions Leslie, 1985 . The DVC promotes ana-bolic activities related to the restoration and con-servation of bodily energy and the resting of vitalorgans. The DVC regulates digestion by modulat-ing digestive polypeptides and gastric motilityŽ .Rogers and Hermann, 1992 . In addition,

Ž .Uvnas-Moberg 1989, 1994 has proposed a paral-lel between DVC regulation of gastrointestinalhormones and the regulation of visceral states,including stress, hunger and satiety. Without ex-ternal challenges, the DVC optimizes the func-tion of the internal viscera. In contrast, by in-creasing metabolic output to deal directly withexternal challenges, the SNS attempts to optimizethe organism’s relationship with the environment.Thus, increases in ambient temperature, noise,pain and pyrogenic agents produce not only in-creased sympathetic activity, but also an active

Žinhibition of DVC actions on the gut Uvnas-Mo-.berg, 1987 .

The paraventricular nucleus of the hypothala-mus is an important regulator of the DVC. Neu-ral communication between the paraventricular


nucleus and the DVC is involved in responsesthat are not only homeostatic, but also protective

Žand defensive e.g. nausea and vomiting, condi-. Žtioned taste aversion, behavioral defense Lawes,

.1990 . Communication between the paraventricu-lar nucleus and the DVC changes with experi-ence, and thus may exhibit a type of learning ormemory. Associations may be established rapidlybetween environmental features or experiencesand visceromotor responses. Perhaps, as in condi-tioned taste aversion, this memory is expressed asa learned association between a specific environ-mental feature and nausea. Once the associationis made, subsequent exposure to the environmen-tal feature may result in immediate nausea anddefensive avoidance behaviors. These specula-tions regarding the changing communicationbetween the paraventricular nucleus and the DVCwith experience are consistent with general theo-

Ž .ries of aversion learning Garcia et al., 1985 .The paraventricular nucleus regulation of the

DVC evolved in phylogenetically older verte-brates, in which escape and avoidance behaviorscontributed to the maintenance of visceral

Ž .homeostasis Lawes, 1990 . Because the early ver-tebrates lacked an elaborate nervous system tocontrol their viscera, behavior was a primarymechanism for the maintenance of homeostasisŽe.g. moving to regulate thermoregulatory and

.oxygen requirements . As the nervous systemevolved, an autonomic nervous system and neu-roendocrine mechanisms emerged and displacedthe need to use behavior to regulate internalstate. The neural and neuroendocrine regulationof internal state allowed behavioral processes tobe directed toward environmental challenges.However, the brain structures, specifically theparaventricular nucleus, that govern thehomeostatically driven behaviors in the phyloge-netically older species evolved into the structuresresponsible for regulating internal homeostaticfunctions in the phylogenetically newer speciesŽ .Leslie et al., 1992 .

The role of the paraventricular nucleus in theregulation of the DVC in modern vertebratesretains phylogenetically older functions and cont-inues to respond to life-threatening situations bycontributing to visceral and endocrine responses

Fig. 4. Neural and neuropeptide regulation of the dorsal vagalŽ .complex DVC in a life-threatening environment. During

life-threatening events when fight�flight behaviors are not anoption, immobilized fear responses are elicited. Immobilizedfear is fostered by vagal surges from the DMX to visceralorgans, which are potentiated by systemic AVP. Systemic AVPtriggers increased DMX output by stimulating visceral affer-ents via NTS and area postrema.

Ž .see Fig. 4 . However, this phylogenetic organiza-tion results in vulnerabilities, because perceivedchallenges to survival, whether or not truly life-threatening, may elicit visceral and endocrine re-actions that compromise normal physiologicalfunction.

The phylogenetic approach of the polyvagaltheory emphasizes that defense and avoidancebehaviors have a vagal component manifestedthrough the DVC. For example, a physiologicalshutdown mediated by the DVC would supportavoidance behaviors, such as feigning death or

Ž .freezing see Fig. 4 . However, the evolution ofhypothalamic regulation of the DVC provides re-sponse alternatives. Specifically, in mammals, theparaventricular nucleus produces two neuropep-tides, oxytocin and vasopressin, that differentiallycommunicate with the sensory and motor por-tions of the DVC. Using the push�pull perfusion


Ž .technique, Landgraf et al. 1990 demonstratedthat both oxytocin and vasopressin are released inthe DVC. Binding sites for vasopressin are preva-lent in the sensory component, but are not repre-

Ž .sented in the motor component Fuxe et al., 1994 .In contrast, oxytocin appears to provide a primarypathway from the paraventricular nucleus to thedorsal motor nucleus of the vagus, with oxytocininjections into the DVC mimicking the vagal re-sponses normally observed immediately following

Ž .feeding Rogers and Hermann, 1992 . Directpathways from the nucleus of the solitary tract tothe paraventricular nucleus provide a potentialsource of feedback for hypothalamic influences

Žon visceromotor functions Sawchenko and Swan-.son, 1982 . Communication between the DVC

and the paraventricular nucleus appears to modu-late specific visceromotor reflexes involving car-

Ž .diovascular Nissen et al., 1993 and gastrointesti-Ž .nal systems Bray, 1985 .

13. A phylogenetic approach to the study ofemotion, stress, and social behavior

The phylogenetic orientation focuses our inter-est on the neural structures and neurobehavioralsystems that we share with, or have adapted from,our phylogenetic ancestry. First, the three re-sponse systems proposed in the polyvagal theoryŽi.e. the cranial nerve regulation of the striatedmuscles of the face coordinated with a myelinatedvagus that inhibits sympathetic activity at thelevel of the heart, a sympathetic-adrenal systemto increase metabolic output, and an inhibitoryvagal system to decrease metabolic output and

.promote freezing and immobilization behaviorsare the products of distinct neurophysiologicalsystems. Second, these distinct neurophysiologicalsystems represent a phylogenetically dependenthierarchy, with the use of cranial nerves to regu-late facial expression emerging in mammalsŽ .well-developed in primates , the sympathetic-adrenal system shared with other vertebrates, in-cluding reptiles, and the inhibitory vagal systemshared with more primitive vertebrates, including

Žamphibians, and bony and cartilaginous fish see.Porges, 1997, 1998 . The three systems, which

have been frequently studied in emotion re-search, represent different phylogenetic stages ofneural development. This phylogenetic develop-ment starts with a primitive behavioral inhibitionsystem, progresses to a fight�flight system, and in

Ž .humans and other primates , culminates in acomplex facial gesture and vocalization system.Thus, from a phylogenetic perspective, the ner-vous system of vertebrates evolved to support agreater range of behaviors and physiologicalstates, including states that we often associatewith stress, as well as positive social behavior.

Even if we focus on stress, the physiologicalresponses of mammals do not fit concisely into asingle neurophysiological system. Although corti-sol has often been labeled as the ‘stress’ hor-mone, other neurally mediated systems clearlyrespond to stress. Moreover, there are situationsof severe stress in which cortisol is neither re-sponsive, nor at a high level. By expanding thephylogenetic model proposed in the polyvagaltheory to include the HPA axis, the hyporespon-sive HPA can be interpreted as reflecting a primi-tive passive avoidance system. The view that theadrenal cortex is the sole stress system at theperiphery limits scientific inquiry into the mecha-nisms and adaptive functions of several interac-tive and complex stress response patterns. Othersystems, including the neuropeptides oxytocin andvasopressin, influence the regulation ofhomeostasis, growth and restoration, metabolismand mobilization. These systems form a complexset of mutually interacting neurophysiologicalpathways that communicate via nerves and neu-

Žrally active chemicals e.g. neurotransmitters,.neuropeptides, hormones to cope with survival

challenges. This paper has attempted to employphylogeny as an organizing principle to illustratethe interactive and adaptive nature of severalphysiological systems that respond to stress andcontribute to the expression of positive socialbehavior.

14. Clinical applications of the polyvagal theory

The polyvagal theory, by describing both thephylogenetically based hierarchy of autonomic


states and the specific triggers that cause a disso-lution of this hierarchy, provides a new way ofinvestigating atypical behavior. The theory em-phasizes that the mammalian nervous system isnot only sensitive to environmental demands andperceived stresses and threats, but that the mam-malian nervous system will, in a predictable or-der, also rapidly reorganize to different neural-mediated states. Unlike models of psychopharma-cology that also emphasize the importance ofstate in regulating behavior, the polyvagal theoryemphasizes neural mechanisms that regulatestate, and the order in which these neural mecha-nisms are elicited. The polyvagal theory empha-sizes the flexibility of the nervous system in modu-lating autonomic state. This view contrasts withpharmacological strategies that are effective inchanging autonomic state without engaging thenervous system.

A new paradigm based on the polyvagal theoryhas been tested. This new paradigm assumes thatstate can be changed in a predictable manner andthat specific state changes would be associatedwith potentiating or limiting the range of specificbehaviors. This new paradigm distinguishesbetween state changes associated with shifts inthe neural regulation of the autonomic nervoussystem and specific responses elicited by quanti-

Ž .fiable stimuli. The paradigm enables: 1 the ma-nipulation and evaluation of autonomic state; andŽ .2 the manipulation and evaluation of learning,social behavior, and other mental processes. Thisparadigm should enable the demonstration that

Žthe range of observable behaviors including phys-.iological reactions will change in a predictable

Ž .manner as the context i.e. perceived safety driv-ing the neural regulation of the autonomic ner-vous system is manipulated. By applying thisparadigm, we can evaluate the influence of speci-fic contexts on neural regulation of the autonomicstate. Some contexts may promote states related

Ž .to antisocial behavior i.e. aggression and healthvulnerability, while other contexts may promotestates related to positive social behavior, healthand development. Thus, unless the design fea-tures of the nervous system are understood withinthe context of the organism’s interaction with theenvironment, an individual may be interacting in

an environment that is not only inefficient instimulating social behavior and intellectualgrowth, but may also be harmful to both mentaland physical health.

The polyvagal theory forces us to interpretcompromised social behavior from a differentperspective. The theory emphasizes that the rangeof social behavior is limited by physiological state.The theory also emphasizes that mobilization andimmobilization behaviors may be adaptive strate-

Ž .gies to a challenged i.e. frightened individual.Thus, it may be possible that creating states ofcalmness and exercising the neural regulation ofbrainstem structures may potentiate positive so-cial behavior by stimulating and exercising theneural regulation of the social engagement sys-tem. This perspective or intervention paradigm,which focuses on biologically based behaviors,might be viewed as a fourth paradigm or ap-proach to modify behavior in contrast to behav-

Ž . Žioral i.e. learning theory-based , biochemical i.e.. Žpharmacological , and psychotherapeutic i.e. in-

.cluding psychoanalysis and cognitive therapiesintervention strategies.

We developed a biologically based behavioralintervention that uses acoustic stimulation to im-prove social behavior, and tested the approachwith children diagnosed with autism. The inter-vention was based on several principles derivedfrom the polyvagal theory. First, the area of the

Ž .brainstem nucleus ambiguus that regulates theŽ .heart i.e. via the myelinated vagus also regulates

the muscles of the head, including those of theface, middle ear, mouth, larynx, and pharynx.Collectively, these muscles function as an inte-grated social engagement system that controlslooking, listening, vocalizing, and facial gesturing.If the neural regulation of this group of muscles

Žis dysfunctional, the face will not work e.g. lackof facial expressiveness, eyelid opening, prosody

.and listening . Interestingly, these facial featuresreflect common behavioral symptoms that havebeen used to describe several psychopathologiesŽe.g. autism, depression, aggressive disorders, and

.posttraumatic stress disorders , emotional statesŽduring severe challenge e.g. grief, rage, anger,

. Žloneliness or medical illness e.g. senility, AIDS,.fever .


Second, the middle-ear muscles play an impor-tant role in extracting human voice from ourcomplex acoustic environment. When neural toneto the middle-ear muscles is low, the middle-earstructures do not actively filter out the low-frequency sounds that dominate the acoustic en-vironment of our modern industrial world and donot amplify the frequencies associated with thehuman voice. This difficulty in listening to thehuman voice might even occur in an individual

Žwho has normal hearing i.e. normal function ofthe cochlea, auditory nerve, and the brain areas

.processing acoustic information .Third, neural regulation of the middle-ear mus-

cles is linked to that of the other muscles of theface, which control facial expression and vocalintonation. Thus, stimulation to improve neuralregulation of the middle-ear muscles should inte-grate and stimulate neural regulation of facialexpression, looking, listening and vocalizing.

The area of the brain that contains the lowermotor neurons for the social engagement systemis near the lower portion of the brainstem. Dur-ing periods of appropriate social communicationŽ .e.g. facial expressiveness, vocal intonation , thelower motor neurons are regulated by the uppermotor neurons in the frontal cortex. During peri-ods characterized by fight�flight behaviors andfear-induced shut-down or immobilization, thetheory proposes that cortical regulation of theselower motor neurons is displaced by phylogeneti-cally more primitive systems. The more primitivesystems are dependent on subcortical structuresthat evolved to negotiate survival by managing

Žmetabolic resources to promote mobilization i.e..fight�flight behaviors or to conserve metabolic

Žresources by immobilizing e.g. freezing or feign-.ing death . Thus, a critical feature of whether an

individual appropriately communicates with thesocial environment, or engages in a strategy offight�flight or freezing behaviors, is determinedby the individual’s perception of the environment.Does the individual perceive the environment assafe or dangerous? The theory states that there isa degrading of the function of the social engage-ment system when the individual perceives theenvironment as dangerous. Conversely, if the in-dividual perceives the environment as safe, there

is the neurophysiological possibility that the cor-tex could regulate the lower motor neurons of thesocial engagement system to promote communi-cation and social behavior. Thus, the perceptionof safety is the primary requirement for our inter-vention.

The intervention was designed to recruit speci-fically cortical regulation of the social engage-ment system to promote the voluntary prosocialbehaviors that are lacking in autistic children.The model is optimistic, because it assumes thatfor many children with social behavior and com-munication difficulties, the social engagement sys-tem is neuroanatomically and neurophysiologi-cally intact. The problem is conceptualized as afunctional deficit. To obtain the desired behavior,the intervention must stimulate the cortical regu-lation of the brainstem system that regulates themuscles of the head. The theory predicts thatonce the cortical regulation of this brainstemsystem is engaged, social behavior and communi-cation will spontaneously occur as the naturalemergent properties of this biological system.Thus, the intervention is seen as stimulation and

Žexercise of a corticobulbar neural system i.e..nerves that connect the cortex to the brainstem

that regulates the muscles of the head.The intervention is based on two primary prin-

Ž .ciples: 1 cortical control of listening and theother components of the social engagement sys-tem via the corticobulbar pathways requires the

Ž .environment to be perceived as safe; and 2exposure to acoustic stimulation within the fre-quency band of the human voice is capable ofstimulating and exercising the neural regulationof the middle-ear muscles and other componentsof the social engagement system. Thus, the inter-vention attempts to engage the active corticalcontrol of the middle-ear muscles as a portal tothe social engagement system.

Our model emphasizes the importance of ex-tracting the human voice in social settings viacortical neural regulation of the middle-ear mus-cles. The intervention uses a relatively narrowfrequency band that focuses on frequencies of thehuman voice. The acoustic stimuli are computer-altered by applying digital filters to extract speci-fic frequencies. The filters are part of a complex


algorithm that modulates the acoustic frequencyband and energy to exercise and stimulate theneural regulation of the middle-ear muscles and,hypothetically, to integrate the components of thesocial engagement system. We focus on the hu-man voice because processing of the human voiceis neurobiologically different from the processingof other acoustic signals, and according to ourmodel, processing of the human voice is a compo-nent of a human social communication system,the social engagement system.

Our initial research protocol applied the inter-vention to children between the ages of 3 and 5years who had a diagnosis of autism. Computer-altered acoustic stimulation was presented in a45-min session, during which attempts were madeto maintain the child in a calm behavioral state.The intervention consisted of five sessions pre-sented on sequential days. Across the five ses-sions, the width of the frequency band and thetotal acoustic energy being presented to the childvaried.

We have tested more than 65 children in adouble-blind, randomized control experimentaldesign. Most children experienced noticeable im-provements in social behavior and communica-tion skills immediately following the intervention.In addition, there were noticeable changes inparental interaction styles with the children, withthe parents being less intrusive following the in-tervention. The improvements in the children’sbehavior persisted when assessed during a 3-month follow-up. We believe that although theintervention was initially tested with children di-agnosed with autism, it will be useful for otherpopulations with difficulties in listening, commu-nicating and organizing social behavior.

15. Conclusions

The polyvagal theory provides a theoreticalplatform to interpret social behavior within aneurophysiological context. The emphasis on phy-logeny provides an organizing principle to under-stand the hierarchical sequence of adaptive re-sponses. The social engagement system not onlyprovides direct social contact with others, but also

modulates physiological state to support positivesocial behavior by exerting an inhibitory effect onthe sympathetic nervous system. From the polyva-gal theory perspective, social behavior is an emer-gent property of the phylogenetic development ofthe autonomic nervous system. Consistent withthis hierarchical model, perceived challenges tosurvival often result in a neural dissolution fromthe more recent systems of positive social behav-ior and social communication to the more primi-tive fight�flight and avoidance systems. The the-ory leads not only to the explanation of thepathophysiological states associated with variousclinical disorders, but also supports the introduc-tion of a new paradigm that may have generalapplications for individuals with difficulties in so-cial behavior.

Acknowledgements

This manuscript is based on an invited keynoteaddress delivered at the 10th World Congress ofthe International Organization of Psychophysi-ology, which was held in Sydney, Australia, inFebruary 2000. The preparation of this manuscriptwas supported in part by grants HD 22628 andMH60625 from the National Institutes of Healthand Human Development. Several of the ideasand concepts described in this paper were theproduct of discussions with C. Sue Carter duringa residency at the Rockefeller Foundation Con-ference Center in Bellagio, Italy.

References

Andrews, P.L.R., Lawes, I.N.C., 1992. A protective role forvagal afferents: a hypothesis. In: Ritter, S., Ritter, R.C.,

Ž .Barnes, C.D. Eds. , Neuroanatomy and Physiology of Ab-dominal Vagal Afferents. CRC, Boca Raton, FL, pp.280�302.

Arima, H., Kondo, K., Murase, T. et al., 1998. Regulation ofvasopressin synthesis and release by area postrema in rats.Endocrinology 139, 1481�1486.

Bazhenova, O.V., Plonskaia, O., Porges, S.W. Vagal reactivityand affective adjustment in infants during interaction chal-lenges. Child Dev., in press.

Braga, D.C., Mori, E., Higa, K.T., Morris, M., Michelini, L.C.,2000. Central oxytocin modulates exercise-induced tachy-cardia. Am. J. Physiol. 278, R1474�R1482.


Bray, G.A., 1985. Autonomic and endocrine factors in theregulation of food intake. Brain Res. Bull. 9, 279�286.

Bueno, L., Gue, M., Fargeas, M.J, . Alvinerie, M., Junien, J.L.,Fioramonti, J., 1989. Vagally mediated inhibition of acous-tic stress-induced cortisol release by orally administeredkappa-opiod substances in dogs. Endocrinology 124,1788�1793.

Buwalda, B., Koolhaas, J.M., Bohus, B., 1992. Behavioral andcardiac responses to mild stress in young and aged rats:effects of amphetamine and vasopressin. Physiol. Behav.51, 211�216.

Cacioppo, J.T., Malaarkey, W.B., Kiecolt-Glaser, J.K. et al.,1995. Heterogeneity in neuroendocrine and immune re-sponses to brief psychological stressors as a function ofautonomic cardiac activation. Psychosom. Med. 57, 154�164.

Cannon, W.B., 1928. The mechanism of emotional disturbanceof bodily functions. N. Engl. J. Med. 198, 877�884.

Cannon, W.B., 1957. ‘Voodoo’ death. Psychosom. Med. 19,182�190.

Ž .reprinted from American Anthropologist 44, 169 1942 .Carpenter, D.O., 1990. Neural mechanisms of emesis. Can. J.

Physiol. Pharmacol. 68, 230�236.Carter, C.S., 1998. Neuroendocrine perspectives on social at-

tachment and love. Psychoneuroendocrinology 23, 779�818.Carter, C.S., Altemus, M., 1997. Integrative functions of lacta-

tional hormones in social behavior and stress management.Ann. NY Acad. Sci. 807, 164�174.

Cheng, Z., Powley, T.L., 2000. Nucleus ambiguus projectionsto cardiac ganglia of rat atria: an anterograde tracing study.

Ž .J. Comp. Neurol. 424 4 , 588�606.Cheng, Z., Powley, T.L., Schwaber, J.S., Doye, F.J., 1999.

Projections of the dorsal motor nucleus of the vagus tocardiac ganglia of rat atria: an anterograde tracing study. J.Comp. Neurol. 410, 320�341.

De Bellis, M.D., Chrousos, G.P., Dorn, L.D. et al., 1994.Hypothalamic�pituitary�adrenal axis dysregulation in sexu-ally abused girls. J. Clin. Endocrinol. Metab. 78, 249�255.

Fan, W., Andresen, M.C., 1998. Differential frequency-depen-dent reflex integration of myelinated and non-myelinatedrat aortic baroreceptors. Am. J. Physiol. 44, H632�H640.

Ferguson, A.V., Lowes, V.L., 1994. Functional neural connec-Ž .tions of the area postrema. In: Barraco, I.R.A. Ed. , Nu-

cleus of the Solitary Tract. CRC, Boca Raton, FL, pp.147�157.

Fuxe, K., Agnati, L.F., Covenas, R., Narvaez, J.A., Bunne-mann, B., Bjelke, B., 1994. Volume transmission in tran-smitter peptide costoring neurons in the medulla oblon-

Ž .gata. In: Barraco, I.R.A. Ed. , Nucleus of the SolitaryTract. CRC Press, Boca Raton, FL, pp. 75�89.

Garcia, J., Laister, P.S., Bermudez-Rattoni, F., Deems, D.A.,1985. A general theory of aversion learning. In: Braveman,

Ž .N.S., Bronstein, P. Eds. , Experimental Assessments andClinical Applications of Conditioned Food Aversions. An-nals of the New York Academy of Sciences.

Gunnar, M.R., Porter, F.L., Wolf, C.M., Rigatuso, J., Larson,M.C., 1995. Neonatal stress reactivity: predictions to lateremotional temperament. Child Dev. 66, 1�13.

Hofer, M.A., 1970. Cardiac respiratory function during suddenprolonged immobility in wild rodents. Psychosom. Med. 32,633�647.

Hopkins, D.A., Bieger, D., de Vente, J., Steinbusch, W.M.,1996. Vagal efferent projections: viscerotopy, neurochem-istry, and effects of vagotomy. Prog. Brain Res. 107, 79�96.

Haselton, J.R., Solomon, I.C., Motekaitis, A.M., Kaufman,M.P., 1992. Bronchomotor vagal preganglionic cell bodiesin the dog: an anatomic and functional study. J. Appl.Physiol. 73, 1122�1129.

Jackson, J.H., 1958. Evolution and dissolution of the nervousŽ .system. In: Taylor, J. Ed. , Selected Writings of John

Hughlings Jackson. Stapes Press, London, pp. 45�118.Jansen, L.M.C., Gispen-de Wied, C.C., Kahn, R.S., 2000. Se-

lective impairments in the stress response in schizophrenicpatients. Psychopharmacology 149, 319�325.

Jones, J.F.K., Wang, X.Y., Jordan, D., 1995. Heart rate re-sponses to selective stimulation of cardiac vagal C fibers inanesthetized cats, rats and rabbits. J. Physiol. 489, 203�214.

Koch, K.L., Summy-Long, J., Bingaman, S., Sperry, N., Stern,R.M., 1990. Vasopressin and oxytocin responses to illusoryself-motion and nausea in man. J. Clin. Endocrinol. Metab.71, 1269�1275.

Lama, A., Delpierre, S., Jammes, Y., 1988. The effects ofelectrical stimulation of myelinated and non-myelinatedvagal motor fibres on airway tone in the rabbit and the cat.Respir. Physiol. 74, 265�274.

Landgraf, R., Mallkinson, T., Horn, T., Veale, W.L., Lederis,K., Pittman, Q.J., 1990. Release of vasopressin and oxytocinby paraventricular stimulation in rats. Am. J. Physiol. 258,155�159.

Lawes, I.N.C., 1990. The origin of the vomiting response: aneuroanatomical hypothesis. Can. J. Physiol. Pharmacol.68, 254�259.

Leslie, R.A., 1985. Neuroactive substances in the dorsal vagalcomplex of the medulla oblongata: nucleus of the tractussolitarius, area postrema, and dorsal motor nucleus of thevagus. Neurochem. Int. 7, 191�211.

Leslie, R.A., Reynolds, D.J.M., Lawes, I.N.C., 1992. Centralconnections of the nuclei of the vagus nerve. In: Ritter, S.,

Ž .Ritter, R.C., Barnes, C.D. Eds. , Neuroanatomy and Physi-ology of Abdominal Vagal Afferents. CRC, Boca Raton,FL, pp. 81�98.

MacLean, P.D., 1990. The Triune Brain in Evolution. PlenumPress, New York.

Miao, F. J-P., Janig, W., Green, P.G., Levine, J.D., 1997.Inhibition of bradykinin-induced plasma extravasation pro-duced by noxious cutaneous and visceral stimuli and itsmodulation by vagal activity. J. Neurophysiol. 78,1285�1292.

Michelini, L.C., 1994. Vasopressin in the nucleus tractus soli-tarius: a modulator of baroreceptor reflex control of heartrate. Braz. J. Med. Biol. Res. 27, 1017�1032.

Morris, J.L., Nilsson, S., 1994. The circulatory system. In:Ž .Nilsson, S., Holmgren, S. Eds. , Comparative Physiology

and Evolution of the Autonomic Nervous System. HarwoodAcademic Publishers, Switzerland, pp. 193�246.


Nissen, R., Cunningham, J.R., Renaud, L.P., 1993. Lateralhypothalamic lesions after baroreceptor-evoked inhibition

Ž .of rat supraoptic vasopressin neurones. J. Physiol. Lond.470, 751�766.

Norris, D.O., 1997. Vertebrate Endocrinology. Academic Press,New York.

Paton, J.F., 1998. Pattern of cardiorespiratory afferent conver-gence to solitary tract neurons driven by pulmonary vagal

Ž .C-fiber stimulation in the mouse. J. Neurophysiol. 79 5 ,2365�2373.

Porges, S.W., 1995. Orienting in a defensive world: mam-malian modification of our evolutionary heritage. A polyva-gal theory. Psychophysiology 32, 301�318.

Porges, S.W., 1997. Emotion: an evolutionary by-product ofthe neural regulation of the autonomic nervous system.Ann. NY Acad. Sci. 807, 62�77.

Porges, S.W., 1998. Love: an emergent property of the mam-malian autonomic nervous system. Psychoneuroen-docrinology 23, 837�861.

Porges, S.W., Doussard-Roosevelt, J.A., Portales, A.L.,Greenspan, S.I., 1996. Infant regulation of the vagal ‘brake’predicts child behavior problems: a psychobiology model ofsocial behavior. Dev. Psychobiol. 29, 697�712.

Potter, E.K., McCloskey, D.I., 1986. Effects of hypoxia oncardiac vagal efferent activity and on the action of thevagus nerve at the heart in the dog. J. Autonomic Nerv.Syst. 17, 325�329.

Qu, L., Hay, M., Bishop, V.S., 1997. Administration of AVP tothe area postrema alters response of NTS neurons to

Ž .afferent inputs. Am. J. Physiol. 272 2 Pt. 2 , R519�R525.Reed, S.F., Ohel, G., David, R., Porges, S.W., 1999. A neural

explanation of fetal heart rate patterns: a test of thepolyvagal theory. Dev. Psychobiol. 35, 108�118.

Richter, C.P., 1957. On the phenomenon of sudden death inanimals and man. Psychosom. Med. 19, 191�198.

Richter, D.W., Spyer, K.M., 1990. Cardiorespiratory control.Ž .In: Loewy, A.D., Spyer, K.M. Eds. , Central Regulation of

Autonomic Function. Oxford University Press, New York,pp. 189�207.

Ž .Rogers RC, Hermann, GE 1992 Central regulation of brain-stem gastric vago-vagal control circuits. In: Ritter S, Ritter,

Ž .RC, Barnes CD Eds , Neuroanatomy and physiology ofabdominal vagal afferents. CRC, Boca Raton, Florida, pp.99�134.

Santer, R.M., 1994. Chromaffin systems. In: Nilsson, S., Holm-

Ž .gren, S. Eds. , Comparative Physiology and Evolution ofthe Autonomic Nervous System. Harwood Academic Pub-lishers, Switzerland, pp. 97�117.

Sawchenko, P.E., Swanson, L.W., 1982. The organization ofnoradrenergic pathways from the brainstem to the paraven-tricular and supraoptic nuclei in the rat. Brain Res. Rev. 4,275�325.

Schwaber, J.S., 1986. Neuroanatomical substrates of cardiovas-cular and emotional�autonomic regulation. In: Magro, A.,

Ž .Osswald, W., Reis, D., Vanhoutte, P. Eds. , Central andPeripheral Mechanisms of Cardiovascular Regulation.Plenum Press, New York, pp. 353�384.

Spyer, K.M., Jordan, D., 1987. Electrophysiology of the nu-cleus ambiguus. In: Hainsworth, R., Williams, P.N., Many,

Ž .D.A.G.G. Eds. , Cardiogenic Reflexes: Report of an Inter-national Symposium. Oxford University Press, Oxford, pp.237�249.

Swanson, L.W., Sawchenko, P.E., 1977. Hypothalamic integra-tion: organization of the paraventricular and supraopticnuclei. Annu. Rev. Neurosci. 6, 269�324.

Taylor, E.W., 1992. Nervous control of the heart and car-diorespiratory interactions. In: Hoar, W.S., Randall, D.J.,

Ž .Farrell, A.P. Eds. , Fish Physiology: The CardiovascularSystem, XII. Academic Press, New York, pp. 343�387.

Uvnas-Moberg, K., 1987. Gastrointestinal hormones andpathophysiology of functional gastrointestinal disorders.

Ž .Scand. J. Gastroenterol. 22 128 , 138�146.Uvnas-Moberg, K., 1989. Gastrointestinal hormones in mother

Ž .and infant. Acta Paediatr. Scand. 351 Suppl. , 88�93.Uvnas-Moberg, K., 1994. Role of efferent and afferent vagal

nerve activity during reproduction: integrating function ofoxytocin on metabolism and behaviour. Psychoneuroen-docrinology 19, 687�695.

Uvnas-Moberg, K., 1997. Physiological and endocrine effectsof social contact. Ann. NY Acad. Sci. 807, 146�163.

Vanhoutte, P.M., Levy, M.N., 1979. Cholinergic inhibition ofadrenergic neurotransmission in the cardiovascular system.

Ž .In: Brooks, C.McC., Koizumi, K., Sato, A. Eds. , Integra-tive Functions of the Autonomic Nervous System. TheUniversity of Tokyo Press, Tokyo, pp. 159�176.

Yehuda, R., Teicher, M.H., Trestman, R.K., Levengood, R.A.,Siever, L.J., 1996. Cortisol regulation in posttraumatic stressdisorder and major depression: a chronobiological analysis.Biol. Psychiatry 40, 79�88.

Documents

The polyvagal theory: phylogenetic substrates of a social ... · The polyvagal theory: phylogenetic substrates of a social nervous system Stephen W. Porges ... dle-ear muscles e.g