Читать книгу Diabetic Neuropathy - Friedrich A. Gries - Страница 8

2 Structure and Function of the Nervous System

N. E. Cameron and C. J. Mathias

The Somatic Nervous System

N. E. Cameron

Overview

The somatic nervous system is subdivided into motor and sensory components. The motor division is concerned with control of skeletal muscle contraction, and hence of voluntary movement and of posture and reflexes. The somatosensory division is a collection of receptors, tracts, and nuclei that convey the sensations of light touch, vibration, temperature, and pain (nociception) to the consciousness. It also conveys information about movements and position of the body (proprioception and kinesthesia). Somatosensory receptors are found in the skin, muscles, joints, and viscera. In addition to providing sensation, the somatosensory division has a critical role in motor control, through feedback about muscle length and tension, joint position, velocity of muscle and limb movement, and contact with external surfaces. The basic structure and function of the somatic nervous system has been described in numerous physiology, neurology, and neuroscience texts: the reader is referred to two recent books for further information [1,2]. Diabetes affects both the peripheral and the central nervous system (for discussion of the latter see Chapter 5, page 205-208), although most clinical and scientific interest has focused on the periphery because of the devastating effect on nerve fiber integrity. This section provides an overview of the substrate of the somatic nervous system, with a greater emphasis on peripheral than on central structures.

Fig. 2.1 General organization of the somatosensory system, showing the dorsal column-lemniscal system which mediates touch sensation and proprioception, and the spinothalamic system, which deals with temperature and fast pain information

Central Nervous System Pathways

Somatosensory System

There are two major somatosensory pathways running from the spinal cord to the primary sensory area (area 3 or S1) on the postcentral gyrus of the cerebral cortex (Fig. 2.1). Information about touch and proprioception is carried in the dorsal column medial lemniscus system, whereas temperature and pain information traverses the spinothalamic pathway [3].

In the touch system, fibers from first-order neurons with cell bodies in the dorsal root ganglia traverse the dorsal columns of the spinal cord (Fig. 2.1) to the dorsal column nuclei within the caudal medulla. Of these, the cuneate nucleus receives input from the upper limbs and body whereas the gracile nucleus deals with fibers from the lower limbs. Second-order neurons then project across the midline as the “sensory decussation” and ascend via the medial lemniscus to the ventral posterolateral nucleus of the thalamus. Touch information from the head follows a parallel route, with the first major relay in the principal trigeminal nucleus of the pons, with mid-pontine decussation to reach the ventral posteromedial nucleus of the thalamus via the ventral trigeminal tract. The inputs from the various regions of the body are segregated and aligned such that the pathway is somatotopically organized. This is reflected by the cortical projection (Fig. 2.1), where there is an orderly mapping of the contralateral side of the body on the cortical surface.

Fig. 2.2 Schematic cross-section of spinal cord showing the gray matter and white matter containing the major sensory (right) and motor (left) nerve tracts

Information about temperature and pain (particularly the fast component) is transmitted by the spinothalamic tract. Here, the first synaptic relay is in the gray matter of the spinal cord; projection fibers then decussate in the ventral white commissure to form the contralateral spinothalamic (or anterolateral) tract (Fig. 2.2).These reach the ventral posterolateral nucleus of the thalamus, and some fibers also project to the small thalamic intralaminar nuclei. Parallel pathways transmit information from the face, via the spinal trigeminal nucleus, to the ventral posteromedial thalamus. The pathway depicted in Figure 2.1 is also known as the neospinothalamic pathway, and there are other, phylogenetically older pathways transmitting pain information (particularly the slow, poorly localized, long-lasting component). These are the paleospinothalamic and spinoreticulothalamic pathways, which are polysynaptic and ascend through the reticular formation to nonspecific nuclei in the medial thalamus and intralaminar nuclei. They project to widespread regions of cerebral cortex, rather than being associated with the somatotopic organization of the primary sensory cortex: this may contribute to the poor ability to localize slow pain.

Motor System

There are a number of important fiber tracts (Fig. 2.2) descending from the brain to control the activity of ventral horn motoneurons supplying the skeletal muscles [2]. These can be divided into lateral and ventromedial groups. The lateral pathway comprises corticospinal and rubrospinal tracts and is primarily involved in voluntary movement, particularly of the distal muscles, under direct cortical control. The ventromedial pathways originate in the brainstem, forming reticulospinal, tectospinal and vestibulospinal tracts, which are involved in the control of posture and locomotion.

The most important component of the lateral pathway is the corticospinal tract, which originates primarily from areas 4 (primary motor cortex or MI) and 6 (supplementary motor area) of the frontal lobe, on the precentral gyrus, located across the central sulcus from the primary somatosensory cortex. The motor cortex is somatotopically organized, and axons pass through the int9rnal capsule and course through the midbrain and pons to form a pyramid-shaped tract running down the ventral surface of the medulla. At the junction with the spinal cord, the pyramidal tract decussates. Thus, as in the somatosensory system, the right motor cortex processes information for the left side of the body and vice versa. The axons from the motor cortex then group to form the lateral corticospinal tract and terminate in the dorsolateral and intermediate gray matter region of the ventral horns, where the motoneurons and interneurons that control the distal muscles are located.

The rubrospinal tract is a much smaller component of the lateral pathway and originates in the red (Latin ruber) nucleus of the midbrain. Axons decussate in the pons and join the corticospinal tract in the lateral columns of the spinal cord. The red nucleus itself receives its major input from the motor areas of cerebral cortex, which also give rise to the corticospinal tract. While the rubrospinal tract is important in many mammalian species, in man much of its function has been taken over by the corticospinal pathway.

The ventromedial pathways may be divided functionally into two groups: the tectospinal and vestibulospinal tracts control the posture of head and neck, whereas the pontine and medullary reticulospinal tracts control the posture of the trunk and limb antigravity muscles. The vestibulospinal tract originates in the medullary vestibular nuclei, which are involved with processing sensory activity from the vestibular apparatus of the inner ear. In combination with proprioceptive information about body and neck position, this pathway is importantly involved in maintaining head-body alignment to ensure that the eyes and our image of the world remain stable [4]. The tectospinal tract originates in the superior colliculus of the midbrain. This structure receives direct input from the retina as well as the visual cortex and auditory systems. It is involved with the coordination of head and eye movements and orienting responses towards stimuli [5].

The pontine reticulospinal tract acts to facilitate the antigravity reflexes of the spinal cord to aid the maintenance of a standing posture by promoting extensor activity in the lower limbs and flexor activity in the upper limbs. The medullary reticulospinal tract has an opposite action, to inhibit reflex domination of anti-gravity muscles, thus allowing greater control by lateral pathways. The balance of activity in these reticulospinal tracts is controlled by descending signals from motor cortex [2].

Fig. 2.3 Schematic of motor control showing the major cortical and subcortical regions of the central nervous system. The association and premotor areas of cerebral cortex, along with the basal ganglia, are responsible for planning and initiating voluntary movements. The motor cortex, and its direct connection with the α-motoneurons in the spinal cord, is responsible for sending the appropriate instructions for execution of the movement by the skeletal muscles. The cerebellum provides information about coordination, sequencing, and timing of complex movements. Brainstem mechanisms, along with the vestibular apparatus, superior colliculus, and older areas of the cerebellum, have a major role in the control of posture and gait

Fig. 2.4 Cross-section of the brain showing the location of the major nuclei that contribute to basal ganglia function upper limbs. The medullary reticulospinal tract has an opposite action, to inhibit reflex domination of antigravity muscles, thus allowing greater control by lateral pathways. The balance of activity in these reticulospinal tracts is controlled by descending signals from motor cortex [2].

While cortical areas 4 and 6 comprise the motor cortex, in terms of the control of voluntary movement many other areas of the cerebral cortex are involved as well as important subcortical structures such as the basal ganglia and cerebellum. The overall organization of the motor control system is outlined in Figure 2.3. Thus, the processes of planning, programming, and initiating goal-directed movement involve association and premotor cortical areas, including the anterior frontal lobes and regions of the posterior parietal cortex such as area 5, which gets a direct input from somatosensory cortex, and area 7, which receives connections from higher-order visual cortical areas.

The major subcortical input to area 6 of motor cortex arises from the ventral lateral nucleus of the thalamus, which in turn comes from the basal ganglia (Fig. 2.4). These comprise the caudate nucleus and the putamen (collectively termed the striatum), the globus pallidus, the subthalamus, and the midbrain substantia nigra. These areas form a complex circuit that funnels or focuses activity from widespread areas of association cortex on to area 6, perhaps supplying basic motor programs for the desired action. This cortex → striatum → globus pallidus → thalamus → motor cortex loop is an integral part of the movement initiation process, and the other structures form side loops that modulate this pathway [6].Diseases of basal ganglia result in problems with voluntary movement initiation, involving hypokinesia, as in Parkinson's disease, or hyperkinesias, as in Huntington's disease. In addition to their role in initiation, basal ganglia also modulate posture and muscle tone, abnormalities of which are found in basal ganglia disease. The basal ganglia also have a role in nonmotor behaviors, including cognition and mood [7].

The cerebellum is involved with coordination of the sequence of muscle contractions during a movement. In man, there are three important functional subdivisions of the cerebellum. The anterior lobe (paleocerebellum, spinal cerebellum) and its associated deep cerebellar nuclei (fastigial, interposed, and lateral vestibular nuclei) is concerned primarily with processing of information from muscle, joint, and cutaneous mechanoreceptors. It also receives input pertaining to activity in the motor cortex via pontine nuclei and collaterals of corticospinal fibers. The anterior lobe output provides information to modulate the brainstem nuclei from which the reticulospinal tracts originate, and there is a projection to the red nucleus. These circuits are involved in the control of posture and gait.

The flocculonodular lobe (archicerebellum, vestibular cerebellum) and fastigial nucleus are involved with the coordination of the paraxial muscles associated with balance and equilibrium. The major input comes from the vestibular apparatus, and output goes to the vestibular nuclei and then to the vestibulospinal tracts. There is also an important vestibulo-ocular projection to the external ocular muscles.

The posterior lobe (neocerebellum, cerebral cerebellum) and dentate nucleus have massive reciprocal connections with cerebral cortex, including motor, premotor, sensory, and posterior parietal areas. This cerebellar subdivision is involved with coordination of voluntary movement sequences, particularly ballistic movements that are normally too fast to be under feed-back proprioceptive control. It is responsible for smooth and accurate execution of movements, and shows evidence of the synaptic plasticity necessary for the learning and refinement of complex motor skills [8].

Peripheral Nerve, Receptors, and Spinal Cord

Peripheral Nerve Fiber Types

Peripheral nerve consists of bundles of nerve fibers, generally mixed sensory and motor. Fiber types have been classified in two ways [2]. The first depends on axon diameter, with categorization into groups A, B, and C. The largest axons, which are myelinated, belong to group A. The smallest fibers, which are unmyelinated, belong to group C. The B group contains myelinated axons from autonomic preganglionic neurons, although this classification is rarely used today. The A group is further classified into the subgroups α, β, δ, and γ.

There is also a second classification system for some of the sensory axons, based primarily on conduction velocity, but also on origin and function. This categorization has numerical classes I–IV in descending order of conduction velocity. Because conduction velocity is directly related to axon diameter for myelinated fibers, the two classification systems can be related as shown in Table 2.1. Both terminologies are in common usage, although they were developed independently and do not overlap exactly in terms of fiber categories.

Somatosensory Receptors and Sensation

Somatosensory receptors can be divided into three groups: mechanoreceptors, thermoreceptors, and nociceptors (Table 2.2). Mechanoreceptors respond to deformation of their nerve endings, which contain specialized mechanosensitive ion channels whose gating depends on stretching or changes in tension of the surrounding membrane. The nerve endings of mechanoreceptors are usually associated with specialized nonneural structures that govern their detailed response characteristics, so determining the adequate stimulus. Mechanoreceptors mediate the sensations of light touch, pressure, vibration and flutter, and limb position and movement (kinesthesia). Examples of mechanoreceptors in hairy and hairless (glabrous) skin are shown in Figure 2.5.

Cutaneous mechanoreceptors [9] have punctate receptive fields whose size is determined by the area of nerve terminal branching and associated nonneural tissue. The information and sensation gleaned from these receptors is governed by their degree of adaptation to a constant stimulus. They may be classified into slowly adapting, moderately rapidly adapting, and very rapidly adapting categories. Slowly adapting receptors respond with an increased frequency of action potentials for the duration of a stimulus. Thus, they are able to accurately signal skin indentation and pressure. Merkel's disk receptors and Ruffini's endings fall into this category. Moderately rapidly adapting receptors respond with a burst of action potentials at stimulus onset and comprise the Meissner's corpuscle of glabrous skin and the hair receptor. They are best at signaling the velocity of movement of a stimulus, and are most sensitive to low-frequency (<50 Hz) repetitive stimulation. In this frequency range, a sinusoidal mechanical stimulus will give rise to the sensation of “flutter” where individual waves of the vibration are felt. This contrasts with higher frequencies, which are felt as a true unitary vibration, and this information is transmitted by the most rapidly adapting receptor type, the pacinian corpuscle.

Fig. 2.5 Schematic of hairy and glabrous (hairless) skin, showing the location of various mechanoreceptors. The receptors in glabrous skin are Meissner's corpuscles and Merkel's disks, located in the dermal papillae, and bare nerve endings. In hairy skin, there are hair receptors around the hair shafts. Merkel's disks, and bare nerve endings. Beneath both types of skin, in the subcutaneous region pacinian corpuscles and Ruffini's endings are found. (From [1], with permission)

The different receptor types work together along with hand movements and skin patterns such as fingerprints for shape and texture discrimination [10]. Thus, while the discharge of slowly adapting Merkel's disks is best at encoding the spatial characteristics of a stimulus, the more rapidly adapting receptors provide texture information from vibrations set up by the mechanical interaction between surface and fingerprints as the finger tip is moved across a surface.

The sensitivity of the skin to mechanical stimulation varies widely over the body [1], as can be seen from the results of two-point spatial discrimination tests (Fig. 2.6). Thus, highest sensitivity is noted for the fingers, lips, nose, and toes, whereas the trunk and upper limbs are relatively insensitive. This is a direct reflection of peripheral innervation density, the number of receptors per unit area of skin, and the average size of individual neural receptive fields, which are correspondingly larger in regions of low sensitivity.

The muscle and skeletal mechanoreceptors comprise muscle spindles, joint receptors, and Golgi tendon organs [11,12]. They are known as proprioceptors because they convey information on the position and movement of the limbs. The major receptor is the muscle spindle (Fig. 2.7), formed from specialized skeletal muscle fibers, the intrafusal fibers. Sensory endings contact the midsection of these fibers in an area devoid of contractile machinery. These afferents are stimulated by stretch and signal muscle length and velocity of lengthening. There is also an efferent supply from specialized γ-motoneurons that innervate the contractile ends of the intrafusal fibers. When activated, this causes the ends to contract and in so doing stretches the noncontractile element, including the afferent endings. The function of this efferent system is to regulate the sensitivity of the afferent fibers during active muscle contractions. In the execution of voluntary movements, when activity is supplied to the contractile (extrafusal) fibers of skeletal muscle via the α-motoneurons, there is also modulatory impulse traffic in the γ system – the principle of α-γ coactivation.

Table 2.2 Classification of peripheral nerve afferent and efferent fibers

Receptor type	Name	Function
Mechanoreceptors
Muscle and skeletal	Muscle spindle	Limb position and motion
	Golgi tendon organ	Muscle tension
	Joint receptor	Joint tension and angle
Cutaneous and subcutaneous	Ruffini's ending	Pressure
	Merkel's disk	Pressure
	Meissner's corpuscle	Touch velocity (hairless skin), low frequency vibration (flutter)
	Hair receptors	Tactile (hairy skin)
	Pacinian corpuscle	Touch acceleration, high-frequency vibration
Thermoreceptors	C bare nerve endings	Warm
	Aδ myelinated	Cold
Nociceptors	Aδ myelinated	High pressure, thermal and mechanothermal
	C bare nerve endings	Thermal and mechanothermal, polymodal, tissue damage products

Fig. 2.6 Two-point discrimination thresholds for different regions of the body, measured as the smallest detectable separation distance between the tips of a calibrated compass. Thresholds vary widely over the body, being at their lowest (2mm) for the finger tips and highest for the forearm, legs, and back (40-50 mm). For selected regions, thresholds are proportional to the diameter of the receptive fields of individual afferents (shown in black). (From [1], with permission)

For the other proprioceptors, joint receptors are located in the connective tissue capsule and they respond to stretch of this tissue to signal joint pressure and angle. Golgi organs are found in the tendons and signal stretch resulting from muscle contraction. In terms of sensation, all these receptor types contribute to the sense of limb position and kinesthesia, along with information from cutaneous mechanoreceptors.

There are separate thermoreceptors for warm or cold stimuli, and like the skin mechanoreceptors they have punctate receptive fields, although they are bare nerve endings rather than encapsulated structures [13]. Warmth is mediated by receptors activated by a range of temperatures between approximately 32 °C and 45 °C, the discharge rate being proportional to temperature. Above 45 °C heat pain, rather than warmth, is perceived, due to the activity of thermal nociceptors, and within this range the discharge of warm receptors actually decreases. Cutaneous cold receptors are activated by temperatures from 1 °C to approximately 20 °C below ambient skin temperature, discharge frequency being roughly proportional to temperature difference. A sensory illusion called paradoxical cold occurs when a 45 °C hot stimulus is selectively applied to a cold fiber receptive field. The stimulus is perceived as cold (rather than warm or painfully hot, which would be the sensation when applied diffusely to the skin), and this coincides with an increased receptor discharge at these high temperatures.

Fig. 2.7 Schematic of the muscle spindle. The main components of the spindle are the intrafusal muscle fibers, sensory afferent endings, and γ-motoneuron efferent fibers. The intrafusal fibers are devoid of contractile apparatus in the region of the afferent endings, although the ends are contractile and are innervated by the γ-motoneurons. The sensory endings are responsive to stretch of the intrafusal fibers. Contraction of the ends of the fibers alters spindle sensitivity. (From [l], with permission)

Nociceptors are also bare nerve endings and respond selectively to stimuli that are sufficiently intense that they could damage tissue, and to chemicals released as a result of tissue damage [14]. Thermal nociceptors respond selectively to extreme heat or cold; mechanical nociceptors are activated by strong mechanical stimulation, most effectively by sharp objects. Chemically sensitive, mechanically insensitive nociceptors respond to a variety of agents including K⁺, extremes of pH, and neuroactive substances such as histamine, bradykinin, and prostanoids, as well as various irritants. Polymodal nociceptors respond to combinations of mechanical, thermal, and chemical stimulation.

Pain is an unpleasant sensation triggered by tissue damage, real or potential. The high emotive content lends subjectivity to the experience, and simple stimulation of nociceptors does not necessarily lead to pain sensation under all circumstances: there are descending pathways that influence pain transmission. The physiological activation of nociceptors, though, usually gives rise to pricking, burning, aching, and stinging sensations. When the skin is damaged, the initial sensation is conveyed by Aδ fibers. C fibers are more important for the longer-lasting perception that outlives the stimulus. However, pain can also result from neural damage and changes in neural circuitry, and does not necessarily require activation of nociceptors. This is clearly seen in the case of phantom limb pain after surgical limb amputation. Such neuropathic pain following peripheral nerve injury and degeneration/regeneration can be caused by alterations in the balance of inputs to the spinal cord sensory neurons. Thus, large myelinated neurons may regenerate better than C fibers, so that spinal cord neurons that once had predominantly nociceptive input could now be dominated by Aβ touch fibers, although the central connection would remain appropriate for pain transmission. Such a phenomenon may underlie allodynia, where previously innocuous stimuli become severely painful [2,15].

Sensory Input to the Spinal Cord

The cell bodies of sensory neurons are located in the dorsal root ganglia. These bipolar neurons have a relatively long peripheral axon branch, and those in the dorsal column-medial lemniscal pathway also have an extensive central axonal projection. Thus, the cell bodies have a considerable task to supply nutrients and materials to maintain axonal function. This consideration, combined with the potential effects of diabetes on dorsal root ganglion microenvironment, could contribute to a relative vulnerability of sensory neurons, which would affect both peripheral and central projections. Evidence for involvement of central axons is seen in a reduction of spinal cord cross-sectional area, determined by magnetic resonance imaging in diabetic patients with distal symmetrical polyneuropathy [16].

The architecture of the spinal input reflects the segmental organization of embryonic development. As the embryo grows and expands, the developing skin carries its segmentally derived innervation with it. Thus, a single area of skin, a dermatome, is supplied by axons from a single dorsal root ganglion [1,2]. This dermatomal organization is shown in Figure 2.8. However, peripheral nerves themselves contain axons from several spinal roots, and different nerves can contain axons from the same root, so the relationship between peripheral nerve trunk and dermatome is complex.

In addition to a central projection, there are also local connections in the spinal cord for neurons that project in the dorsal column-medial lemniscal pathway. Second-order neurons for the spinothalamic projection are also located in the dorsal horn of the spinal cord [17]. The synaptic connections for the different sensory fiber types are made in different layers of the dorsal horn. Thus, Aδ fibers synapse primarily in laminae I and V; C fibers connect mainly in laminae II and I; Aβ input goes to lamina V neurons.

Control of Pain Transmission

Pain transmission from the spinal cord can be modified by nonpainful sensory input as well as by activation of descending pathways from various brain nuclei. Painful sensations evoked by activity of nociceptors (Aδ and C fibers) can be decreased by simultaneous stimulation of low-threshold mechanoreceptors, which may underlie the pain-reducing effects of rubbing of the skin and transcutaneous and dorsal column electrical stimulation. This derives from the gate theory of pain[18], according to which pain results from the balance of activity in nociceptive and nonnociceptive afferent fibers. Thus, nonnociceptive Aβ fiber activity “closes” the central transmission gate whereas nociceptive activity “opens” it. The detailed spinal cord circuitry responsible for this effect is not known, but neurons in lamina V receive convergent input from Aβ. Aδ, and C fiber afferents. Furthermore, Aβ fiber activity can suppress firing of lamina V neurons via inhibitory interneurons in lamina II.

Spinal pain transmission is also modulated by descending inputs [19,20]. Neurons in the periaqueductal gray matter of the midbrain make connections with the cells in the rostroventral medulla, particularly in the nucleus raphe magnus. These in turn project to the spinal cord and make inhibitory connections with neurons in laminae I, II, and IV. Thus, electrical stimulation of periaqueductal gray or raphe nuclei inhibits dorsal horn neurons, including those giving rise to the spinothalamic tract, providing profound analgesia. There are several important neurotransmitter systems involved. The periaqueductal gray area has a very high density of opioid receptors. The raphe nucleus contains many serotonergic neurons. Another descending pathway, from the midbrain locus ceruleus, is noradrenergic. The serotonergic and noradrenergic fibers stimulate dorsal horn interneurons that release the endogenous opiate neurotransmitter enkephalin to pre- and postsynaptically inhibit spinothalamic tract projection neurons.

Motor Output and Sensorimotor Integration in the Spinal Cord

The final common motor pathway to the skeletal muscles is via the motoneurons, whose axons make up a substantial proportion of the myelinated fiber population of peripheral nerve. The major motor output of the spinal cord comes from the α-motoneurons, which directly stimulate skeletal muscle force production by synaptic activation at the motor end plate. The other cord output, from γ-motoneurons, exerts an indirect influence on muscle tension by controlling muscle spindle sensitivity and dynamic range, which consequently affects reflex activation of α-motoneurons.

The basic element of motor control is called the motor unit, which comprises an α-motoneuron together with all of the muscle fibers that it innervates [21]. The collection of α-motoneurons that innervates a single skeletal muscle is termed the motor unit pool of that muscle. The size of the motor units varies greatly. For muscles involved in high-precision movements such as those of the digits or face, the number of muscle fibers may be only 10. For large muscles of the trunk and limbs there may be 3000-4000 fibers per motor unit. These fibers may be distributed over the entire area of the muscle, so the territory of a single motor unit may be considerable. All muscle fibers in an individual motor unit are biochemically, histochemically, and physiologically identical, indicating the determination of muscle fiber properties by their innervation. In terms of muscle and contractile properties, there are three motor unit types. Those based on type I muscle fibers are characterized by relatively slow contraction speeds, reliance on aerobic energy metabolism, and a profuse capillary supply-features that confer extreme fatigue resistance. These units are active for much of the time, being involved in postural control, and are preferentially recruited by muscle spindle afferent input to the spinal cord. In contrast, type IIB motor units have fast contraction times, rely on anaerobic metabolism, have relatively poor vascular supply, and fatigue rapidly. They are preferentially recruited for large, fast movements such as limb withdrawal from a painful stimulus. Type IIA units are somewhere in between: the muscle fibers are fast contracting, well supplied with capillaries, have both aerobic and anaerobic energy production, and are moderately fatigue-resistant.

Fig. 2.8 Distribution of dermatomes and peripheral nerve patterns. Mapping of sensory innervation of the skin by the dorsal roots is shown on the left of the subject: cervical (C1-C7), thoracic (T1-T12), lumbar (L1-L5), and sacral (S1-S5). There is no dorsal root at C1, only a ventral (motor) root. The innervation patterns of peripheral nerves are shown for comparison on the left. Individual peripheral nerves have fibers that arise from several adjacent dorsal roots, leading to rather larger fields of innervation and overlap in the area innervated by each segment. (From [2], with permission)

Fig. 2.9 Schematic of the stretch or myotatic spinal reflex circuitry. The essential component is an excitatory (+) monosynaptic connection between muscle spindle la afferent fibers and the α-motoneuron pool for that muscle. The reflex may be evoked transiently by tapping the muscle tendon, which stretches the muscle and spindle endings, causing a short reflex contraction. Also shown is a connection via an inhibitory interneuron (−, black cell body), which suppresses activity in the antagonist muscle

There are two fundamental ways of varying muscle tension production: altering the frequency of action potentials transmitted by an individual a-motoneuron, and altering the range or number of motor units activated in that muscle. This is dependent on the synaptic input to the α-motoneurons, of which there are three sources: muscle spindle afferents, the corticospinal projection, and spinal cord interneurons. The latter category forms a complex control mechanism as it is in turn strongly influenced by both afferent input and all descending motor pathways.

The elegance and simplicity of the spinal cord circuitry involved in sensorimotor integration is apparent for the myotatic or stretch reflex. Thus, Sherrington [22] noted that when a muscle is stretched it tends to contract. This was traced to an excitatory monosynaptic reflex arc between muscle spindle afferents, which are stimulated by the stretch, and the α-motoneurons, which cause that muscle to contract (Fig. 2.9). The operation of this circuit is used as a clinical tool, observing the reflex jerk when the tendons are tapped to rapidly stretch muscle. Physiologically, however, this simple circuit acts tonically as a length servo feedback loop, which is crucial for postural stability and has important antigravity functions, for example in the major leg extensors. This circuit is elaborated by interneurons to ensure that antagonistic muscle groups controlling the same joint do not work against each other [23]. Thus, collateral branches of the spindle afferents also synapse on inhibitory interneurons that in turn innervate the motor unit pool of the antagonist muscle.

Another spinal circuit involving proprioceptor input is the reverse myotatic (or clasp knife) reflex arc (Fig. 2.10). Golgi tendon organs, which signal muscle tension, are the sensory arm and they innervate inhibitory interneurons that synapse with the α-motoneurons [1,2]. Thus, the reflex is polysynaptic, with increasing muscle tension tending to inhibit further contraction. This may be demonstrated, for example, when a subject is asked to actively resist bending of their knee. At a high level of applied force, the strongly contracted extensors suddenly relax and the resistance disappears. This reflex arc, in extreme circumstances, may function to protect muscles from inappropriately high and potentially damaging tension production. However, the circuitry forms a physiological tension servo feedback mechanism to maintain a particular contracted state of the whole muscle, for example when some fibers weaken or drop out due to fatigue. It also is important for steady tension production required in fine motor control, for example in holding a fragile object. Thus, the tension servo mechanism acts in conjunction with the length servo mechanism provided by the muscle spindles to maintain precise positioning of limbs, postural maintenance, and manipulation of grip.

A third important spinal circuit involves nociceptive input and the complex polysynaptic reflex arc that mediates flexion or withdrawal of a limb from an aversive stimulus [1,2], for example standing on a nail (Fig. 2.11). Pain fibers enter the spinal cord and branch profusely to activate excitatory interneurons, which in turn excite α-motoneurons to flexor muscles. The magnitude of the noxious stimulus governs the size of the withdrawal response and the number of flexors responding: a highly painful stimulus will excite all the flexors of the affected limb. Thus, this reflex crosses dermatomal boundaries and involves integration between several spinal segments. As with the myotatic reflex, this circuit is elaborated to incorporate reciprocal inhibition of the antagonistic extensor muscles. Furthermore, a postural component is added, the crossed-extensor reflex, to support the weight of the body on the contralateral leg when the foot is with-drawn.

Fig. 2.10 Circuits for the inverse myotatic or clasp knife spinal reflex. This uses information from the lb afferents of Golgi tendon organs, which monitor tendon stretch and muscle tension. The reflex may be evoked by attempting to stretch a muscle during an isometric contraction. This causes a rapid increase in tension and a massive volley of action potentials from the tendon organs. Via inhibitory interneurons (–, black cell body), α-motoneuron activity is suppressed and the limb collapses, similar to the closing of a clasp knife blade

From this brief description, it is clear that the spinal cord carries out the basic steps of sensorimotor integration, which are further elaborated by the brain. However, the spinal cord circuits are not there simply to carry out reflex actions to sensory stimulation. The same circuits are recruited by descending inputs from the brain to simplify voluntary movement and postural adjustment. For example, the crossed-extensor reflex pathway (one leg flexed, the contralateral leg extended, and vice versa) is also an element in the sequencing of walking.

Vascular Supply in the Peripheral Somatic Nervous System

Blood flow to the nervous system is influenced or regulated by a number of factors including local tissue metabolism, oxygen and carbon dioxide tension, pH, circulating vasoactive agents, the intrinsic innervation of the blood vessels, and systemic perfusion pressure. This has been extensively documented for the cerebral circulation (see reviews [24–26]). However, the precise details of vascular regulation differ in the central and peripheral nervous system: moreover, the vascular supply has different characteristics in dorsal root ganglia and nerve trunks of the somatosensory nervous system. Given the importance of impaired blood flow in several peripheral nerve disease states, including diabetic neuropathy (see Chapter 4, page 115-123), a brief overview of the salient features of the normal vascular supply to the peripheral somatosensory system is appropriate.

Fig. 2.11 Circuits for the flexion (withdrawal) and crossed-extension reflexes. These reflexes are mediated by polysynaptic pathways in the spinal cord. Noxious stimulation of Aδ fibers causes excitation of α-motoneurons supplying the ipsilateral flexor muscles, which withdraw the limb from the threat of damage. Excitatory interneurons also connect to the α-motoneurons of the extensors on the contralateral limb, causing contraction to support the weight of the body during limb withdrawal. α-Motoneurons supplying the antagonistic muscles on both sides of the cord are inactivated via inhibitory interneurons during this reflex

Peripheral Nerve Trunk and Spinal Roots

The vasculature of peripheral nerve is relatively unique. Peripheral nerve and its spinal dorsal and ventral roots have a good vascular supply composed of two integrated but independent systems, termed the extrinsic and intrinsic circulations [27–29]. The extrinsic system comprises vessels that arise from local large arteries and veins as well as offshoots from the vessels supplying adjacent muscles and periosteum. These are arranged segmentally and follow the surface along the length of the nerve (Fig. 2.12). They form a highly anastomotic plexus within the epiperineurial layers of nerve sheath, vessels being mainly arterioles, venules, and arteriovenous shunts. This provides numerous connections with the intrinsic circulation. The latter, or vasa nervorum, comprises vessels on the perineurium and in the endoneurial vascular bed. Terminal arterioles from the perineurium penetrate the nerve fascicles and form the endoneurial capillary bed, which consists of a network of intrafascicular capillaries that run longitudinally, along with the nerve fibers, throughout the length of the nerve. Endoneurial capillaries are lined by a continuous layer of endothelial cells connected by tight junctions, which forms part of the blood-nerve barrier, analogous to, although somewhat less efficient than, the blood-brain barrier that restricts ingress of blood-borne substances to the central nervous system. The other component of the blood-nerve barrier is the inner layer of the perineurium, which is continuous with the arachnoid membrane of spinal cord [30].

Fig. 2.12 Peripheral nerve gross structure and blood supply. Peripheral nerves are surrounded by a loose connective tissue structure, the epineurium, which contains a plexus of blood vessels supplied by radial branches from multiple feed arteries. Nerves are divided into fascicles by the perineurium, which is a strong connective tissue that isolates the fascicles of nerve fibers physically to form part of the blood-nerve barrier. The other component of the barrier is the tight endothelial lining of the endoneurial capillaries

The endoneurial capillaries have a large diameter compared to those in other tissues such as brain and skeletal muscle, and the intercapillary distance is relatively great [31,32]. The latter would tend to render nerve susceptible to ischemic or hypovolemic stresses or nerve edema [33]. However, the extensive anastomotic connections between extrinsic and intrinsic circulatory systems minimize the effect of local disruptions to nerve blood supply, and, coupled with a low metabolic rate, make peripheral nerve relatively resistant to mild ischemia. The intrinsic circulation consists predominantly of capillaries, and there is a paucity of vascular smooth muscle in the endoneurium [32]. Thus, the main neural and humoral control of nerve perfusion is exerted at the level of the epiperineurial arterioles. These vessels are densely innervated by nerve fibers [34,35] containing a number of neurotransmitters, including norepinephrine, serotonin, substance P, neuropeptide Y, vasoactive intestinal peptide, and calcitonin gene-related peptide. By contrast, although nerves containing these neurotransmitters are also found in endoneurium, they are not associated with blood vessels.

The capacity for autoregulation—to match blood flow to local tissue demand—is well established in the central nervous system. Myogenic mechanisms compensate for changes in perfusion pressure to maintain tissue flow. Levels of metabolites such as O₂, CO₂, and pH influence vessels to supply blood to meet metabolic needs. However, in peripheral nerve, pressure autoregulation of endoneurial blood flow is virtually absent, perhaps a reflection of the lack of smooth muscle in endoneurial microvessels. Furthermore, vasa nervorum responses to systemic hypoxia, hypercapnia, or reduced pH are minimal [36]. These factors would make peripheral nerve more vulnerable than many tissues to hypotension-induced ischemia. Conversely, there is some evidence of local metabolite regulation. Following repetitive electrical nerve stimulation, a functional hyperemia is apparent [37]. Moreover, blood flow can be seen to approximately double during maximal but selective electrical stimulation of large myelinated nerve fibers, which do not innervate vasa nervorum [38].

Peripheral Ganglia

Dorsal root (and autonomic) ganglia have blood flow values three to five times greater than the blood flow of peripheral nerve, which reflects the higher metabolic rate of neuronal cell bodies compared to that of nerve fibers [39–41]. The capacity of ganglia to autoregulate their blood supply is considerably better than that of peripheral nerve. Thus, ganglia show excellent flow regulation in the face of changes in systemic blood pressure. However, in common with peripheral nerve and in contrast to brain, there is little response to changes in systemic arterial CO₂ or pH [39].

Although dorsal root ganglia are surrounded by an impermeable perineurium, a proportion of the vessels have a fenestrated endothelial lining. This renders the ganglion-blood barrier weak, and dorsal root ganglion cells are vulnerable to blood-borne toxic substances such as heavy metals, some anticancer drugs, and infectious agents. The reason for this breakdown of the blood-nervous system barrier is not known, but it has been suggested that a subpopulation of ganglion cells may be chemical sensors providing important information on the body's internal milieu [42].

Concluding Remarks

The somatic nervous system, with its motor and sensory divisions, forms the basic output and monitoring capability for all movement, both purposive and reflex. The elaborate organization of proprioceptors and skin mechanoreceptors and their central connections are essential for our sense of body position in space, the control of posture, gait, and accuracy of goal-directed movement. The skin is a particularly important sensory organ, the sentry at the interface of body and environment, monitoring not only mechanical but also thermal and potentially damaging events.

The Autonomic Nervous System

C.J. Mathias

Introduction

The autonomic nervous system innervates every organ in the body and is closely involved in their function (Fig. 2.13). Additionally, it plays a key role in integrative function and in maintaining the milieu intérieur, for example through the control of blood pressure, body temperature, and metabolic and fluid balance. This enables optimum functioning in a variety of situations, which at times is essential for survival. It is accomplished by numerous pathways and neurotransmitters, providing considerable flexibility and responsiveness. Dysfunction of the autonomic nervous system may be caused at one or more sites centrally or peripherally (Table 23). It can be a particular problem in diabetes mellitus, where neural structures can be affected at various sites and their disturbances compounded by target organ involvement. In this section the principles of structure and function of the autonomic nervous system will be described, along with examples pertaining to derangement of function.

Basic Principles

The autonomic nervous system is essentially an efferent system encompassing sympathetic, parasympathetic, and enteric components. The sympathetic efferents emerge from the thoracic and lumbar segments of the spinal cord and ultimately supplyall organs and structures. The parasympathetic outflow consists of cranial and sacral efferents. The former accompany cranial nerves III, VII, IX, and X, and supply the eye, lachrymal and salivary glands, heart, lungs, and upper gastrointestinal tract with associated structures down to the level of the colon. The sacral outflow supplies the large bowel, urinary tract, bladder, and reproductive system. The enteric nervous system as originally proposed by Langley in 1898 is effectively the local nervous system of the gut.

There are specific cerebral nucleii, especially in the hypothalamus, midbrain, and brainstem, that influence autonomic activity; an example is the Edinger-Westphal nucleus through the parasympathetic nerves to the iris musculature which controls pupillary constriction. From the brainstem, sympathetic efferent outflow tracts descend through the cervical spinal cord, where axons synapse in the intermediolateral cell mass (Fig. 2.14). From the thoracic and upper lumbar spinal segments, myelinated axons emerge and synapse in the paravertebral ganglia, which is some distance from sympathetically innervated target organs. Most parasympathetic ganglia are close to target organs. The major neurotransmitter at ganglia (both parasympathetic and sympathetic) is acetylcholine, which acts on the nicotinic receptor (Fig. 2.15). Postganglionic fibers, which are unmyelinated, rejoin the mixed nerves through the gray rami and innervate target organs except for the adrenal medulla, which only has a preganglionic supply. The neurotransmitter at sympathetic postganglionic synapses (as in the heart and blood vessels) is predominantly norepinephrine; sympathetic cholinergic fibers (with acetylcholine as the neurotransmitter that acts on muscarinic receptors) supply sweat glands.

Fig. 2.13 Parasympathetic and sympathetic innervation of major organs. (From [43], with permission)

Table 2.3 Classification of disorders resulting in autonomic dysfunction^a

PRIMARY (Etiology unknown) Acute/subacute dysautonomias Pure cholinergic dysautonomia Pure pandysautonomia Pandysautonomia with neurological features Chronic autonomic failure syndromes Pure autonomic failure Multiple system atrophy (Shy-Drager syndrome) Parkinson's disease with autonomic failure SECONDARY Congenital Nerve growth factor deficiency Hereditary Autosomal dominant trait Familial amyloid polyneuropathy Porphyria Autosomal recessive trait Familial dysautonomia (Riley-Day syndrome) Dopamine β-hydroxylase deficiency Aromatic L-amino acid decarboxylase deficiency X-linked recessive Fabry's disease Metabolic diseases Diabetes mellitus Chronic renal failure Chronic liver disease Vitamin B12 deficiency Alcohol-induced Inflammatory Guillain-Barré syndrome Transverse myelitis Infections Bacterial: tetanus Viral: human immunodeficiency virus Infection Parasitic: Trypanosoma cruzi (Chagas' disease) Prion: fatal familial insomnia Neoplasia Brain tumors: especially of third ventricle or posterior fossa Paraneoplastic, to include adenocarcinomas of lung, pancreas, and Lambert-Eaton syndrome Connective tissue disorders Rheumatoid arthritis Systemic lupus erythematosus Mixed connective tissue disease Surgery Regional sympathectomy: upper limb, splanchnic Vagotomy and drainage procedures: “dumping syndrome” Organ transplantation: heart, kidney Trauma Spinal cord transection DRUGS. TOXINS and POISONS Direct effects By causing a neuropathy (alcohol) NEURALLY MEDIATED SYNCOPE Vasovagal syncope Carotid sinus hypersensitivity Micturition syncope Cough syncope Swallow syncope Associated with glossopharyngeal neuralgia POSTURAL TACHYCARDIA SYNDROME

^aAdapted from [44].

Fig. 2.14 Anatomic components of the central autonomic network. The diagram of the human brain indicates the areas involved in central autonomic control, as defined by animal studies. The insular cortex is the primary viscerosensory area. The central nucleus of the amygdala is involved in emotional responses. The paraventricular, lateral and other hypothalamic regions are involved in homeostasis and adaptive behavior. The periaqueductal gray integrates autonomic, motor, and antinociceptive receptive responses during stress. The parabrachial region is a viscerosensory relay and participates in cardiovascular and respiratory control. The nucleus of the tractus solitarii (nucleus of the solitary tract) is the primary viscerosensory relay nucleus. The ventrolateral medulla contains sympathoexcitatory, sympathoinhibitory, and respiratory neurons. The nucleus ambiguus contributes innervation to the heart. The intermediate reticular zone of the medulla (shaded area) is critically involved in integration of respiratory, cardiovascular, and other autonomic reflexes. (From [45], with permission)

The autonomic nervous system often works as a servo system enabling responsiveness to a variety of local and systemic influences, with interaction and at times control at different levels of the neural axis. Thus, every afferent in the body can influence parasympathetic and sympathetic efferent activity (Fig. 2.16). Examples are visual afferents (through cranial nerve II) that influence pupillary function; chemoreceptors, sinoaortic baroreceptors, and low-pressure receptors through cranial nerves IX and X, which regulate cardiac vagal and sympathetic neural control of heart and blood vessels; and visceral, skin, and muscle receptors (through cerebral connections or via the isolated spinal cord as observed in patients with high spinal cord lesions), which influence a wide range of autonomic activity.

Fig. 2.15 Outline of the major transmitters at autonomic ganglia and postganglionic sites on target organs supplied by the sympathetic and parasympathetic efferent pathways. The acetylcholine receptor at all ganglia is of the nicotinic subtype (ACh-n). Ganglionic blockers such as hexamethonium thus prevent both parasympathetic and sympathetic activation. Atropine, however, acts only on the muscarinic (ACh-m) receptor at postganglionic parasympathetic and sympathetic cholinergic sites. The cotransmitters along with the primary transmitters are also indicated. NA, norepinephrine (noradrenaline); VIP, vasoactive intestinal polypeptide; NPY, neuropeptide Y. (From [44], with permission)

Fig. 2.16 Schema to indicate the major afferent pathways that influence the major autonomic efferent outflow (the cranial and sacral parasympathetic and the thoracolumbar sympathetic), supplying various organs. Cr = Cranial nerve (From [44], with permission)

Fig. 2.17 Schema of pathways in the formation, release, and metabolism of norepinephrine from sympathetic nerve terminals. Tyrosine is converted into dihydroxyphenylalanine (DOPA) by tyrosine hydroxylase (TH). DOPA is converted Into dopamine (DA) by dopadecarboxylase. In the vesicles DA is converted into norepinephrine (noradrenaline, NA) by dopamine β-hydroxylase (DβH). Nerve impulses release both DβH and NA into the synaptic cleft by exocytosis. NA acts predominantly on α₁-adrenoceptors but has actions on β-adrenoceptors on the effector cell of target organs. It also has presynaptic adrenoceptor effects. Those acting on α₂-adrenoceptors inhibit NA release: those on β-adrenoceptors stimulate NA release. NA may be taken up by a neuronal process (uptake 1) into the cytosol, where it may inhibit further formation of DOPA through the rate-limiting enzyme TH. NA may be taken into vesicles or metabolized by monoamine oxidase (MAO) in the mitochondria. NA may be taken up by a higher-capacity but lower-affinity extraneuronal process (uptake 2) into peripheral tissues, such as vascular and cardiac muscle and certain glands. NA is also metabolized by catechol-o-methyl transferase (COMT). NA measured in plasma is the overspill not affected by these numerous processes. (From [44], with permission)

There is increasing evidence in humans, resulting from a combination of neuroendocrine and neuroimaging studies, of the role of various cerebral areas that control, influence, and modulate autonomic function. This is in keeping with extensive experimental data of the role of the insular cortex and amygdala in cardiovascular responses, and of hypothalamic nucleii in neuroendocrine control.

In the enteric nervous system, prevertebral ganglia (celiac, superior and inferior mesenteric) have both sympathetic and parasympathetic efferents and a system of neurons and supporting cells within various viscera including the gastrointestinal tract, pancreas, and gall bladder. These innervate the musculature of the alimentary tract (and thus influence gut motility), the secretion of organs (such as the flow of gastric acid), mucosal blood flow, and the intestinal transport of water and electrolytes. There are sensory neurons that monitor factors such as tension in the walls of the intestine or the chemical nature of its content, and associated neurons (interneurons) that link information between enteric and motor neurons and influence smooth muscle contraction, vasodilatation, and transport of water and electrolytes. They interact with sympathetic and parasympathetic pathways and the wide range of enteric endocrine cells, with numerous pancreatic and gut peptides that have roles both locally and elsewhere. Within the enteric nervous system are a number of interconnected networks or plexuses. In the intestines these include the myenteric (Auerbach's) plexus between the external longitudinal and circular muscle coats, and the subserous (Meissner's) plexus in the connective tissue between the serosal mesothelium and external muscle. The major neurotransmitter is acetylcholine, and these plexuses are thus similar to the intrinsic plexuses found in the heart that also are the sites of selective involvement in Chagas' disease following infection with Trypanosoma cruzi; dysfunction may result from a specifically targeted immunological process.

Fig. 2.18 Schema of pathways in the formation from choline of acetylcholine (ACh) and its inactivation by acetylcholine esterase (AChE). (From [46], with permission)

The postganglionic autonomic supply to organs and effector cells consists of multiple neurotransmitters with complex machinery relating to their formation, release, interplay with other substances, uptake, and recycling. Schema for the threemajor neurotransmitters, norepinephrine in adrenergic, acetylcholine in cholinergic, and adenosine triphosphate in purinergic terminals, are provided in Figures 2.17-2.19. A variety of other substances may be cosecreted, examples being vasoactive intestinal polypeptide with acetylcholine, and neuropeptide Y with norepinephrine (Fig. 2.15). This may explain the inability of specific antagonists to block all the effects of parasympathetic and sympathetic nerve stimulation, as observed with the muscarinic blocker atropine and with α-adrenoceptor blockers, respectively.

Fig. 2.19 Purinergic junction: scheme of synthesis, storage, release, and inactivation of autonomic transmitters. (From [47], with permission)

Assessing Autonomic Activity

Activity of the autonomic nervous system can be evaluated directly or indirectly. In the sympathetic nervous system this can be performed by electrophysiological (with sympathetic microneurography) or biochemical techniques. The latter include measurement of catecholamines and their metabolites in urine and plasma, spillover techniques using tritiated catecholamines, and use of substances such as metaiodobenzylguanidine (MIBG, a γ-emitter) or 6-[¹⁸F] fluorodopamine (a positron emitter) that are taken up by postganglionic sympathetic nerves and detected by appropriate scanning techniques. However, there are limitations to each of these techniques. Sympathetic microneurography is dependent upon insertion of a fine tungsten microelectrode into a peripheral nerve and into a muscle or skin sympathetic fascicle (Fig. 2.20). It is an invasive technique, although safe in experienced hands; there are difficulties with electrode placement in autonomic disorders causing underactivity or inactivity. Urinary and plasma catecholamines (Fig. 2.21) do not control for small or rapid changes and are affected by metabolism and uptake effects, among other factors. The spillover techniques are invasiveand involve direct cannulation of blood vessels, but provide measurement of change in sympathetic activity, especially in key organs such as the heart, kidneys, and brain (Fig. 2.22). The imaging techniques using MIBG or fluorodopamine are noninvasive but semiquantitative. These different techniques are used extensively in research but not usually in the routine clinical setting, where tests of function are utilized that are dependent not only on activity of the autonomic nerves but also on the response of target organs.

Fig. 2.20 Relationship between spontaneous fluctuations of blood pressure and muscle nerve sympathetic activity recorded in the right peroneal nerve. Arterial baroreflex activity accounts for the pulse synchrony of nerve activity and the inverse relationship to blood pressure fluctuations. The asterisk indicates a diastolic blood pressure fall due to sudden atrioventricular block. Stippling indicates corresponding sequences of bursts and heart beats. (From [48], with permission)

Investigation of Autonomic Function

An outline of investigational approaches for relevant systems is provided in Table 2.4. A detailed history and clinical examination, to include all organs and integrative system function affected by autonomic dysfunction, helps guide the choice of testing. The effects of autonomic underactivity and overactivity should be considered. Cardiovascular autonomic assessment often provides a readily assessable and noninvasive means of screening for dysfunction. A variety of cardiovascular tests are helpful in distinguishing between sympathetic and parasympathetic function (Figs. 2.23, 2.24). A cardinal feature of sympathetic denervation often is orthostatic (postural) hypotension (Fig. 2.25), which will not be detected unless measurements are made in the supine and head-up (sitting or standing) positions. These measurements often are performed when there is clinical suspicion, such as when syncope is reported. However, there can be a variety of symptoms resulting from, or in association with, orthostatic hypotension (Table 2.5) that have multiple causes and this may result in failure to consider orthostatic hypotension and thus measure blood pressure before and after postural challenge. Orthostatic hypotension may occur later in the course of disease, as in diabetes mellitus, where cardiac parasympathetic denervation often is anearly feature of autonomic neuropathy. When orthostatic hypotension is present, consideration also should be given to a range of causative nonneurogenic factors (Table 2.6).

Table 2.4 Outline of investigations in autonomic failure

Cardiovascular
Physiological	Head-up tilt (45°); standing; Valsalva maneuverPressor stimuli: Isometric exercise, cold pressor, mental arithmeticHeart rate responses: deep breathing, hyperventilation, standing, head-up tilt, 30:15 ratioLiquid meal challengeExercise testingCarotid sinus massage
Biochemical	Plasma norepinephrine: supine and head-up tilt or standing; urinary catecholamines; plasma renin activity and aldosterone
Pharmacological	Norepinephrine: a-adrenoceptors, vascularIsoprenaline: (β-adrenoceptors, vascular and cardiacTyramine: pressor and norepinephrine responseEdrophonium: norepinephrine responseAtropine: parasympathetic cardiac blockade
Sudomotor	Central regulation: thermoregulatory sweat testSweat gland response: intradermal acetylcholine, quantitative sudomotor axon reflex test (Q-SART).localized sweat testSympathetic skin response
Gastrointestinal	Barium studies, video-cine-fluoroscopy, endoscopy, gastric emptying studies
Renal function and urinary tract	Day and night urine volumes and sodium/potassium excretionUrodynamlc studies, intravenous urography, ultrasound examination, sphincter electromyography
Sexual function	Penile plethysmographyIntracavernosal papaverine
Respiratory	LaryngoscopySleep studies to assess apnea/oxygen desaturation
Eye	Lachrymal function: Schirmer's testPupillary function: pharmacological and physiological

Adapted from [49]

A variety of other tests related to stimuli in daily life need to be considered in relation to diagnosis, understanding pathophysiological mechanisms, and management. Examples include the effects of food and exercise among other factors (Table 2.7) that unmask or exaggerate orthostatic hypotension when there is sympathetic vasoconstrictor failure. Food can causea marked fall in blood pressure because of splanchnic vasodilatation and an inability to compensate in other vascular regions: exercise causes vasodilatation in working muscles. Ambulatory 24-hour blood pressure (Fig. 2.26) and heart rate profiles are of value as lack of the expected nocturnal circadian fall indicates autonomic failure; with suitable protocols, orthostatic, postprandial, and exercise-induced hypotension can be evaluated with these ambulatory techniques in the home setting. Spectral analytical techniques provide further information on the differences between sympathetic and parasympathetic control of heart rate and blood pressure, and can assess respiratory influences over heart rate, in particular.

Sudomotor testing should include evaluation of gustatory sweating when relevant. In combination with neurophysiological tests, the sympathetic skin response is of value in determining sympathetic cholinergic activation (Fig. 2.27). Details of tests affecting other systems can be obtained from various textbooks [46,52,53].

Evaluation of Central Autonomic Activity and Function

The evaluation of central autonomic activity and function is separately described as there are difficulties in accurately making measurements noninvasively. Recently, however, various technological and analytical advances have been utilized to advantage. Neuroimaging using widely available techniques such as brain magnetic resonance imaging is repeatable and reproducible and can determine morphology of even small structures such as the insular cortex, amygdala, and pontine regions; thus, in central autonomic disorders such as multiple system atrophy, discrete abnormalities are discernible in the brainstem. Further amplification of neuronal involvement may be obtained by magnetic resonance spectroscopy, although abnormalities have been described mainly in the basal ganglia. Of importance are the techniques of positron emission tomography (PET) and functional MRI (fMRI) scanning. In normal humans, specific areas may be activated by different stimuli, such as the anterior cingulate gyrus for cardiovascular tasks (Fig. 2.28), and the amygdala, with varying hemispheric dominance in response to different emotional stimuli (Fig. 2.29). The use of various neuropsychological paradigms to stimulate different brain areas, especially the amygdala, should be of further value in determining the functional anatomy of cerebral autonomic centers in normal humans, and in evaluating disturbances of function in various autonomic disorders.

Fig. 2.21 Plasma norepinephrine, epinephrine, and dopamine levels (measured by high-pressure liquid chromatography) innormal subjects (controls), patients with multiple system atrophy (MSA) or pure autonomic failure (PAF) and two individual patients with dopamine β-hydroxylase deficiency (DBH defn) while supine and after head-up tilt to 45° for 10 minutes. The asterisks indicate levels below the detection limits for the assay, which are less than 5 pg/ml for norepinephrine and epinephrine and less than 20 pg/ml for dopamine. Bars indicate ± SEM. (From [49], with permission)

Fig. 2.22 Pre-and postprandial regional plasma norepinephrine spillover, indicating sympathetic nervous system activation in normal subjects, a The open histograms indicate the values while fasting and the filled histograms the postprandial values, in different vascular beds. The percentage changes are indicated in b. There is greater activation in the renal and skeletal muscle vasculature than in cardiac or hepatic regions. (From [50], with permission)

Table 2.5 Some symptoms resulting from orthostatic hypotension and impaired perfusion of various organs

Cerebral hypoperfusion Dizziness Visual disturbances Blurred vision Tunnel vision Scotoma Graying out Blacking out Color defects Loss of consciousness Impaired cognition Muscle hypoperfusion Paracervical and suboccipital (“coathanger”) ache Lower back/buttock ache Calf claudication Cardiac hypoperfusion Angina pectoris Spinal cord hypoperfusion Renal hypoperfusion Oliguria Nonspecific Weakness, lethargy, fatigue Falls

Adapted from [44]

Fig. 2.23 The effect of deep breathing on heart rate and blood pressure in a a normal subject and b a patient with autonomic failure. There is no sinus arrhythmia in the patient, despite a fall in blood pressure. Respiratory changes are indicated in the middle panel. (From [49], with permission)

Fig. 2.24 Changes in intra-arterial blood pressure before, during, and after the Valsalva maneuver, when intrathoracic pressure was raised to 40 mmHg in a normal subject (upper trace) and in a patient with autonomic failure (lower trace). In the normal subject, release of intrathoracic pressure was accompanied by an increase in blood pressure and reduction in heart rate below basal levels. In the patient there was a gradual increase in blood pressure, implying impairment of sympathetic vasoconstrictor pathways. The heart rate scale varies in the two subjects. (From [49], with permission)

Fig. 2.25 Blood pressure and heart rate before, during, and after head-up tilt in a normal subject (uppermost panel), a patient with autonomic failure (middle panel), and a patient with vasovagal syncope (lowermost panel). In the normal subject there is no fall in blood pressure during head-up tilt, unlike the patient with autonomic failure, in whom blood pressure falls promptly and remains low with a blood pressure overshoot on return to the horizontal. In the patient with autonomic failure there is only a minimal change in heart rate despite the marked blood pressure fall. In the patient with vasovagal syncope there was initially no fall in blood pressure during head-up tilt; in the latter part of tilt, as indicated in the record, blood pressure initially rose and then fell to low levels, so that the patient had to be returned to the horizontal. Heart rate also fell. In each case continuous blood pressure and heart rate was recorded with the Portapress II. (From [49], with permission)

Table 2.6 Nonneurogenic causes of orthostatic hypotension

Low intravascular volume
Blood/plasma loss	Hemorrhage, bums, hemodialysis
Fluid/electrolyte	Inadequate intake: anorexia nervosa
	Fluid loss: vomiting, diarrhea, losses from ileostomy
	Renal/endocrine: salt-losing neuropathy, adrenal insufficiency (Addison's disease).
	diabetes insipidus, diuretics
Vasodilatation
	Drugs: glyceryl trinitrate
	Alcohol
	Heat, pyrexia
	Hyperbradykininism
	Systemic mastocytosis
	Extensive varicose veins
Cardiac impairment
Myocardial	Myocarditis
Impaired ventricular filling	Atrial myxoma, constrictive pericarditis
Impaired output	Aortic stenosis

Adapted from [49]

Another approach to central autonomic evaluation has been the use of neuroendocrine challenge tests, where physiological or pharmacological stimuli in conjunction with measurement of neuroendocrine markers provide, in vivo, an indication of which controlling nucleus, pathway, or even neurotransmitter is involved. This has been demonstrated in multiple system atrophy (MSA), where the lesions are predominantly central, as compared to pure autonomic failure (PAF), where they are peripheral. Thus, the argininevasopressin (AVP) response to a physiological stimulus causing baroreceptor activation (with head-up tilt) is abnormal in both MSA and PAF; however, central osmoreceptor stimulation (with hypertonic saline and rise in AVP), is preserved in PAF, but not in MSA. An example of a neuroendocrine test utilizing a neuropharmacological stimulus is based on the α₂-adrenoceptor projection to the hypothalamus, which causes release of human growth hormone releasing factor (HGRF), that then raises growth hormone (GH) levels. In normal subjects, central stimulation with the α₂-agonist clonidine elevates HGRF and GH; this also occurs when there is peripheral autonomic denervation without central involvement, as in PAF (Fig. 2.30). However, in MSA, with predominantly central lesions, there is no GH response to clonidine. This lack of rise in GH is not due to an inability of hypothalamic cells to secrete HGRF (and thus stimulate the pituitary), as the GH secretagogue L-dopa, raises both HGRF and GH levels in MSA (Fig. 2.31). This favors a specific abnormality in the α₂-adrenoceptor projection to the hypothalamus in MSA, and indicates that the impaired GH response to clonidine is not due to hypothalamic neuronal cell loss or pituitary malfunction. These neuroendocrine mapping approaches therefore help to determine both the central sites and the neurotransmitters involved, especially in autonomic disorders that may selectively affect one or more neuronal systems and pathways.

Table 2.7 Factors influencing orthostatic hypotension

Speed of positional change

Time of day (worse in the morning)

Prolonged recumbency

Warm environment (hot weather, central heating, hot bath)

Raising intrathoracic pressure: micturition, defecation, or coughing

Food and alcohol ingestion

Physical exertion

Physical maneuvers and positions (bending forward, abdominal compression, leg crossing, squatting, activating calf muscle pump)a

Drugs with vasoactive properties (including dopaminergic agents)

Adapted from [49]

^a These maneuvers usually reduce the postural fall in blood pressure, unlike the others.

Fig. 2.26 Twenty-four-hour noninvasive ambulatory blood pressure and heart rate profiles showing systolic (•—•) and diastolic (•···•) blood pressure and heart rate at intervals through the day and night, a Changes in a normal subject with no postural fall in blood pressure; there was a fall in blood pressure at night whilst asleep (an expected circadian nocturnal fail), with a rise in blood pressure on wakening, b Marked fluctuations in blood pressure in a patient with autonomic failure. The falls in blood pressure are usually the result of postural changes, either sitting or standing. Blood pressure when supine, particularly at night, is elevated. Getting up to micturate causes a marked fall in blood pressure (0300 hours). There is a reversal of the diurnal changes in blood pressure. The changes in heart rate are relatively small, considering the large changes in blood pressure. (From [49], with permission)

Fig. 2.27 a The sympathetic skin response (in microvolts) from the right hand and right foot of a normal subject (control) and a patient with dopamine (β-hydroxylase (DBH) deficiency. In pure autonomic failure and pure cholinergic dysautonomia (b) the sympathetic skin response could not be recorded. (From [51], with permission)

Additional Nonautonomic Investigations

In clinical practice, evaluation of autonomic function and dysfunction often needs to be combined with other investigations to determine the causative or associated disease (as in secondary autonomic disorders), and whether there are coexistent diseases or complications. This is of particular importance in diabetes mellitus, where multiple systems may be affected.

Thus investigation of both large and small cerebral vessels (to exclude cerebrovascular disease), neurophysiological assessment (to determine the presence and extent of motor and sensory neuropathy), and allied investigations into urinary bladder, gut, and sexual function may be needed to distinguish between neurogenic failure of target organs and other causes of organ dysfunction (see [44,46,52,53]).

Fig. 2.28 Right anterior cingulate activity showing positive covariance with mean arterial blood pressure (MAP) during isometric exercise and mental arithmetic tasks. Activity in the right anterior cingulate (ac) covaried significantly with increasing blood pressure. For all subjects, regional activity covarying with MAP was computed for isometric exercise and mental arithmetic tasks. (From [54], with permission)

Fig. 2.29 Views of the brain showing activation of the left amygdala using PET scanning with H₂¹⁵0 superimposed on structural magnetic resonance images, and the construction of a statistical parametric map. This was in response to visual stimuli using photographs showing happy and fearful facial expressions. Regional cerebral blood flow (rCBF) is indicated on the right. Faces with fearful expressions caused greater change in blood flow to the amygdala than happy faces. (From [55], with permission)

Fig. 2.30 The upper panel (a) shows mean (SE) serum growth hormone (GH) concentrations before (0) and at 15-minute intervals for 60 minutes after administration of clonidine (2 µg/kg per minute) in normal subjects (controls) and in patients with pure autonomic failure (PAF) and multiple system atrophy (MSA). GH concentrations rise in controls and in patients with PAF with a peripheral lesion; there is no rise in patients with MSA with a central lesion. The lower panel (b) Indicates lack of serum GH response to clonidine In the two forms of MSA (the cerebellar form, MSA-C, and the parkinsonian form, MSA-P), in contrast to patients with idiopathic Parkinson's disease (IPD) with no autonomic deficit, in whom there is a significant rise in GH levels. (From [56], with permission)

Fig. 2.31 Mean (SE) serum growth hormone (GH), plasma human growth hormone releasing hormone (GHRH), and dopamine concentrations before and after administration of l-dopa in nine patients with MSA-P. *P< 0.05 and †P< 0.01 vs basal (time 0). (From [56], with permission)

References

[1] Kandel ER, Schwartz JH, Jessell TM, editors. Principles of neural science. 4th ed. New York: McGraw-Hill: 2000.

[2] Kingsley RE. Concise text of neuroscience, 2nd ed. Baltimore: Lippincott. Williams & Wilkins; 2000.

[3] Gardiner EP. Martin JH, Jessell TM. The bodily senses. In: Kandel ER, Schwartz JH, Jessell TM, editors. Principles of neural science, 4th ed. New York: McGraw-Hill; 2000:430-50.

[4] Keshner EA, Cohen H. Current concepts of the vestibular system reviewed: 1. The role of the vestibulospinal system in postural control. Am J Occup Ther 1989; 43:320-30.

[5] Wurtz RH, Albano JE. Visual-motor function of the primate superior colliculus. Annu Rev Neurosci 1980; 3:189-226.

[6] Brooks DJ. The role of the basal ganglia in motor control: contributions from PET. J Neurol Sci 1995; 128:1-13.

[7] Stocchi F, Brusa L. Cognition and emotion in different stages and subtypes of Parkinson's disease. J Neurol 2000; 247 (Suppl 2):II114-21.

[8] Thach WT. A role for the cerebellum in learning movement coordination. Neurobiol Learning Memory 1998; 70:177-88.

[9] Birder LA, Perl ER. Cutaneous sensory receptors. J Clin Neurophysiol 1994; 11:534-52.

[10] Johnson KO, Hsiao SS. Neural mechanism of tactual form and texture perception. Annu Rev Neurosci 1992; 15:227-50.

[11] Hulliger M. The mammalian muscle spindle and its central control. Rev Physiol Biochem Pharmacol 1984; 101:1-110.

[12] Zimny ML. Mechanoreceptors in articular tissues. Am J Anat 1988; 182:16-32.

[13] Spray DC. Cutaneous temperature receptors. Annu Rev Physiol 1986; 48:625-38.

[14] Cesare P, McNaughton P. Peripheral pain mechanisms. Curr Opinion Neurobiol 1997; 7:493-9.

[15] Baron R. Peripheral neuropathic pain: from mechanisms to symptoms. Clin J Pain 2000: 16 (Suppl):S12-S20.

[16] Eaton SE, Harris ND, Rajbhandari SM, Greenwood P, Wilkinson ID, Ward JD, Griffiths PD, Tesfaye S. Spinal-cord involvement in diabetic peripheral neuropathy. Lancet 2001; 358:35-6.

[17] Fields HL. Pain. New York: McGraw-Hill; 1987.

[18] Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965; 150:971-9.

[19] Bausbaum AI, Fields HL. Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry. Annu Rev Neurosci 1984; 7:309-38.

[20] Willis WD, Westlund KN. Neuroanatomy of the pain system and of the pathways that modulate pain. J Clin Neurophysiol 1997; 14:2-31.

[21] Burke RE. Motor unit properties and selective involvement in movement. Exerc Sport Sci Rev 1975; 3:31-81.

[22] Liddell EG, Sherrington C. Reflexes in response to stretch (myotatic reflexes). Proc R Soc Lond B Biol Sci 1924; 96:212-42.

[23] Lundberg A. Multisensory control of spinal reflex pathways. Prog Brain Res 1979; 50:11-28.

[24] Paulson OB, Strandgaard S, Edvinsson L. Cerebral autoregulation. Cerebrovasc Brain Metab Rev 1990; 2:161-92

[25] Szabo C. Physiological and pathophysiological roles of nitric oxide in the central nervous system. Brain Res Bull 1996; 41:131-41.

[26] Faraci FM, Heistad DD. Regulation of the cerebral circulation: role of endothelium and potassium channels. Physiol Rev 1998; 78:53-97.

[27] Lundborg G, Branemark PI. Microvascular structure and function of peripheral nerves. Adv Microcirc 1:66-88.

[28] Lundborg G. Intraneural microcirculation. Orthop Clin North Am 1988; 19:1-12.

[29] Petterson CA, Olsson Y, Blood supply of spinal nerve roots. An experimental study in the rat. Acta Neuropathol 1989; 78:455-61.

[30] Olsson Y. Microenvironment of the peripheral nervous system under normal and pathological conditions. Crit Rev Neurobiol 1990; 5:265-311.

[31] Bell MA, Weddell AG. A descriptive study of the blood vessels of the sciatic nerve in the rat, man and other mammals. Brain 1984; 107:871-98.

[32] Bell MA, Weddell AG. A morphometric study of intrafascicular vessels of the mammalian sciatic nerve. Muscle Nerve 1984; 7:524-34.

[33] Low PA, Lagerlund TD. McManis PG. Nerve blood flow and oxygen delivery in normal, diabetic and ischemic neuropathy. Int Rev Neurobiol 1989; 31:355-438.

[34] Appenzeller O, Dhital KK, Cowen T, Bumstock G. The nerves to blood vessels supplying blood to nerves: the innervation of vasa nervosum. Brain Res 1984; 304:383-6.

[35] Rechthand E, Hervonen A, Sato S, Rapoport SI. Distribution of adrenergic innervation of blood vessels in peripheral nerve. Brain Res 1986; 374:185-9.

[36] Low PA, Tuck RR. Effects of changes of blood pressure, respiratory acidosis and hypoxia on blood flow in the sciatic nerve of the rat. J Physiol (Lond) 1984; 347:513-24.

[37] Kihara M, Nickander KK, Low PA. The effect of aging on endoneurial blood flow, hyperaemic response and oxygen free radicals in rat sciatic nerve. Brain Res 1991; 562:1-5.

[38] Cameron NE, Cotter MA, Robertson S, Maxfield EK. Nerve function in experimental diabetes in rats: effects of electrical stimulation. Am J Physiol 1993; 264:E161-6.

[39] McManis PG, Schmelzer JD, Zollman PJ, Low PA. Blood flow and autoregulation in somatic and autonomic ganglia: comparison with sciatic nerve. Brain 1997; 120:445-9.

[40] Sasaki H, Schmelzer JD, Zollman PJ, Low PA. Neuropathology and blood flow of nerve, spinal roots and dorsal root ganglia in longstanding diabetic rats. Acta Neuropathol 1997; 93:118-28.

[41] Cameron NE, Cotter MA. Diabetes causes an early reduction in autonomic ganglion blood flow in rats. J Diabetes Complications 2001; 15:198-202.

[42] Devor M. Unexplained peculiarities of the dorsal root ganglion. Pain 1999; Suppl 6:S27-35.

[43] Janig W. Autonomic nervous system. In: Schmidt RF, Thews G, editors. Human physiology, 2nd ed. Berlin: Springer-Verlag; 1987: 333-70.

[44] Mathias CJ. Disorders of the autonomic nervous system. In: Bradley WG, Daroff RB, Fenichel GM, Marsden CD, editors. Neurology in clinical practice. 3rd ed. Boston: Butterworth-Heinemann; 2000: 2131-65.

[45] Westmoreland BF, Benarroch EE, Daube JR, et al. Medical neurosciences: an approach to anatomy, pathology and physiology by systems and levels. 3rd ed. Boston: Little, Brown and Co: 1994.

[46] Appenzeller O, Oribe E. The autonomic nervous system, 5th ed. Amsterdam: Elsevier; 1997.

[47] Bumstock G. Autonomic neuroeffector functions - reflex vasodilatation of the skin. J Invest Derm 1997; 69:47-57.

[48] Wallin BG, Linblad L-E. Baroreflex mechanisms controlling sympathetic outflow to muscles. In: Sleight P, editor. Arterial baroreceptors and hypertension. Oxford: Oxford University Press 1980:101.

[49] Mathias CJ, Bannister R. Investigation of autonomic disorders. In: Mathias CJ, Bannister R, editors. Autonomic failure: a textbook of clinical disorders of the autonomic nervous system. 4th ed. Oxford: Oxford University Press; 1999: 169-95.

[50] Vaz M, Cox HS, Kaye DM, Turner AG, Jennings GL, Esler MD. Fallibility of plasma noradrenaline measurements in studying postprandial sympathetic nervous responses. J Auton Nerv Syst 1995; 56: 97-104.

[51] Magnifico F, Misra VP, Murray NMF, Mathias CJ. The sympathetic skin response in peripheral autonomic failure-evaluation in pure autonomic failure, pure cholinergic dysautonomia and dopamine beta-hydroxylase deficiency. Clin Auton Res 1998; 8: 133-38.

[52] Low PA., editor. Clinical autonomic disorders, 2nd ed. Philadelphia: Lippincott-Raven; 1997.

[53] Mathias CJ, Bannister R, editors. Autonomic failure: a textbook of clinical disorders of the autonomic nervous system. 4th ed. Oxford: Oxford University Press; 1999.

[54] Critchley HD, Corfield DR, Chandler MP, Mathias CJ, Dolan RJ. Cerebral correlates of autonomic cardiovascular arousal: a functional neuroimaging investigation in humans. J Physiol 2000; 523: 259-70.

[55] Morris JS, Frith CD, Perreth DI, et al. A differential neural response in the human amygdala to fearful and happy facial expressions. Nature 1996; 383: 812-5.

[56] Kimber JR, Watson L, Mathias CJ. Distinction of idiopathic Parkinson's disease from multiple system atrophy by stimulation of growth hormone release with clonidine. Lancet 1997; 349: 1877-81.