Chapter 4

SPECIFIC FEARS, VAGUE ANXIETIES AND THE AUTONOMIC NERVOUS SYSTEM

A. INTRODUCTION

B. EXPERIMENTAL CONDITIONS FOR THE STUDY OF FEAR

Pavlovian Conditioning

Delay conditioning

Trace conditioning

Instrumental Conditioning

Escape learning

Avoidance learning

Two factor theory

Generalized fears

Conditioned emotional response

Punishment

Conflict

Two way avoidance

The Human Condition

C. THE BODY'S RESPONSE TO FEAR AND ANXIETY

The Adrenal Flight or Fight Response

General Adaptation Syndrome

Surgical Shock

Sudden Death

Ulcers

Executive monkeys

The triad design

Control of stressors

Prediction of stressors

Presence of conflict

Stressors Revisited

D. THE PHARMACOLOGY OF STRESS RESPONSES

The Search for Autonomic Stabilizers

The Tranquilizers (Phenothiazines)

The Antianxiety Drugs (Benzodiazepines)

From Laboratory to Clinic and Back

Receptors for Phenothiazines

Receptors for Benzodiazepines

Anticholinergics as Anti-punishment Drugs

Treatment of Ulcers

E. THE AUTONOMIC RESPONSE: CHICKEN OR EGG?

James-Lange Theory of Emotion

Schachter and Singer's Model

F. SUMMARY

Principles

Terms


Return to main Table of Contents>


SPECIFIC FEARS, VAGUE ANXIETIES AND

THE AUTONOMIC NERVOUS SYSTEM

A. INTRODUCTION

Fear is one of the most pervasive and pernicious of human conditions. At low levels, it is a frequent contributor to normal activities, providing both energy and direction to our interactions with the physical and social environments. In its clinical manifestations, fear contributes to a variety of psychiatric disorders including phobias, compulsions, panic attacks, and anxieties both specific and vague. It is the cornerstone for mental distress which leads to a long (and growing) list of psychosomatic disorders such as hypertension, headache, ulcers, cardiovascular disease, and immunological dysfunction--to name a few. We would do well to further our understanding of fear and anxiety. The goal of this chapter is to interrelate: (a) the environmental conditions that lead to various states of fear, (b) the changes in the central and autonomic nervous systems that accompany these environmental conditions, and (c) the pharmacological interventions that are available for research and therapy.

The term fear is firmly entrenched in both our technical vocabulary and the vernacular. It would serve little purpose to offer a precise definition for the present discussion, but some consideration of the ways in which the term has been defined may be useful. For research purposes, conditioned fear typically is defined operationally in terms of the environmental events that set the conditions for fear, although some behavioral or physiological index is frequently used to corroborate the effect. Irrespective of the precision of this operational definition, the ultimate goal of the research (and, indeed, the choice of parameters that make up the operational definition) is to provide an experimental model that parallels the human conditions listed above. This remains a distant goal, but major developments within the areas of experimental psychology and psychopharmacology allow a reasonably coherent model to be presented.

The terms conditioned fear or learned fear underscore the importance of experience in determining the sources of fear. Some theorists claim that all fears are learned during the lifetime of the individual. Even those who argue for innate fears present a very short list of exemplars (e.g, fear of snakes, fear of unsupported heights). In virtually all cases, the fear is based upon some consistent relationship between the environment and some painful or otherwise noxious stimulus. It is the understanding of these environmental relationships that provides us with the experimental procedures to study fear and anxiety.

B. EXPERIMENTAL CONDITIONS FOR THE STUDY OF FEAR

Pavlovian Conditioning

The earliest experiments, and those which have become most important in our understanding of fear, were done by Pavlov (1927). His reasons for conducting these experiments were not to learn about fear and anxiety, but rather to develop the laws for learning about environmental relationships. An important distinction was the one made between the unconditioned response (UR) and the conditioned or conditional response (CR) of the organism. The UR is the direct response that is elicited by the noxious stimulation. Examples put forth by Pavlov include the defensive salivation in response to the sour taste of acid, leg flexion in response to foot shock, and other motor responses to intense physical stimuli such as a pin prick. Pavlov recognized the importance of the psychic (i.e., emotional) component of this direct response to the strong stimulation and, more importantly, the ability of this emotional component to move forward in time and anticipate the occurrence of the painful stimulation. Pavlov studied this phenomenon in considerable detail, but the three paradigms shown in Figure 4-1 demonstrate the most important principles that he developed:

Delay conditioning

The delay conditioning procedure can also utilize long delays with somewhat different results. In this procedure, the CS is gradually presented for longer and longer periods of time until there is eventually a long delay between the onset of the CS and the occurrence of the painful US. Pavlov found that the animals not only could bridge this gap in time, but ultimately were able to appropriately delay the occurrence of the CR until it just preceded the arrival of the US.

Trace conditioning

The trace conditioning procedure involves a brief presentation of the CS, a period of time during which no stimulus is presented, and then the presentation of the painful stimulus. Under these conditions, the anticipatory responding is slower to develop and more fragile, but the success of the procedure provided the necessary demonstration that the conditioned response could be based upon the memory (trace) of a previous stimulus.

In summary, Pavlov developed the experimental procedures to study the important facts that emotional behaviors such as fear do not require the actual presence of an aversive event, but can be triggered (in a lawful fashion) by events that have reliably predicted the occurrence of aversive events. As we shall see later, it is this separation in time of emotional behavior from the actual events that elicit the original response that forms the basis for the development (and treatment) of stress disorders.

Instrumental Conditioning

Escape learning

Pavlov's experimental procedures involved the physical restraint of the subjects, thereby limiting the types of anticipatory responses that could be made. The experimental procedures that evolved in the United States involve much less restraint and allow more global response patterns to be emitted. One of the more common parallels to Pavlov's unconditioned response places the subject (usually a rat rather than a dog) in a long, narrow alleyway that has an electrified grid floor (see Fig. 4-2). The subject can escape from this painful stimulation by running to the opposite end of the alley and stepping into the non-electrified goal area of the box. Note that the subject's behavior is instrumental in escaping from the aversive stimulation. Learning is evidenced by progressively faster running speeds.

Avoidance learning

This simple escape procedure is typically modified to include an initial warning signal (i.e., a CS) that allows a brief period of time to reach the goal area before the shock (US) arrives. Thus, the rat can either avoid the shock by traversing the alley during the presentation of the CS or, failing that, can escape the shock that follows several seconds later. For the philosophically myopic, the ability of the rat to rapidly learn to avoid the impending shock so readily was a problem: Was the rat performing some behavior that was based upon some future set of events? Well, of course it was, but the acceptance of this notion was aided greatly by the proposal of the so-called two factor theory of avoidance behavior.

Two factor theory

Some of the early experimentalists saw the distinction between classical and instrumental conditioning as being too arbitrary, and suggested that instrumental conditioning may include a component of classical conditioning (e.g., Mowrer, 1939). The two factor theory suggests that avoidance behavior is based upon a combination of classical (Pavlovian) conditioning and instrumental conditioning. The subject first learns the environmental relationships that exist according to the laws of Pavlovian conditioning, for example, the onset of a light (CS) is reliably followed ten seconds later by the onset of shock (US) to allow the development of anticipatory fear. (Other stimuli such as handling, the characteristics of the testing chamber, etc., can also serve as CSs). Once this anticipatory fear has been established, the organism can learn the environmental contingencies that are based on the fact that certain responses are instrumental in terminating either the fear-producing CS or the actual pain-producing US. Thus, the notion of conditioned fear becomes an important determinant in the selection of behavior, and the so-called avoidance responses are actually responses that escape this conditioned fear.

Generalized fears

In the simple experimental procedures described above, the conditioned fear plays a straightforward, even positive, role in guiding the behavior of the organism. There are, however, a variety of situations in which the same fear response interferes with ongoing behavior and, as we will see shortly, contributes to the harmful physiological effects of stress. One of the early demonstrations of a learned fear response that is basically nonproductive was Watson and Rayner's (1920) somewhat infamous experiment with Little Albert. A few presentations of a white stuffed toy (CS) followed by a loud noise (US) resulted in a learned fear response that could be elicited by the presentation of the CS alone. In fact, this learned emotional response was elicited not only by the original stuffed toy, but by other similar white furry objects-- a phenomenon called stimulus generalization. Although the procedures and the underlying processes of learning are essentially identical to Pavlov's simple conditioning procedure, the resulting conditioned response seems less adaptive than leg flexion in a restraining harness. The conditioned fear in these types of situations can be maintained for long periods of time, perhaps indefinitely, through interaction with other behaviors. Individuals who have such fears (e.g., phobias) typically adopt behaviors (avoidance responses) that prevent or minimize contact with the fear eliciting stimuli. This not only results in undesirable restriction of unrelated activities, but also allows many situations that are only remotely related to elicit low levels of fear or anxiety in anticipation of approaches to the original stimulus.

Conditioned emotional response

This interference with unrelated behaviors formed the basis for another experimental model which is termed conditioned emotional response (CER) or conditioned suppression (cf., McAllister & McAllister, 1971). In this procedure, some baseline behavior such as lever pressing for food reward is established. After stable rates of responding have been attained, a long-lasting CS (e.g, a 90-sec tone) is presented and terminates with the presentation of a brief intense shock. After a few such pairings of the tone and shock, the subject will suppress responding during the CS presentation, even though the food reward contingency is still in effect, and completely independent of the tone-shock pairings. The usual interpretation of this is that the tone elicits a conditioned fear response which is incompatible with feeding. This procedure has been used in countless experiments as the prototype of situations in which conditioned fear interferes with other, unrelated behaviors. Ironically, the powerful influence of the CER situation can be attributed to the actual lack of relationship to the lever pressing that is food rewarded.

Punishment

The punishment procedure contrasts sharply with the noncontingent shock presentation that characterizes the CER. Punishment procedures specifically deliver an aversive stimulus each time a particular response is made, and have the advantage of greatly narrowing the range of suppressed behaviors, leaving most other behaviors unchanged. Another way of looking at this phenomenon is that the behavior per se comes to serve as the CS which predicts shock. Other behaviors do not predict shock and, hence, do not lead to the learned fear that suppresses ongoing activity.

Conflict

Punishment procedures are not without problems. To the extent that the behavior in question is strongly motivated, the delivery of punishment can lead to a situation of conflict. One of the most widely used conflict procedures which will be referred to in several cases later, is the procedure developed by Geller and Seifter (1960). This procedure combines several elements of the experimental situations that have been described above. First, the subjects are trained to press a lever to obtain some positive reinforcer such as food, which is usually presented on a variable interval (VI) schedule of reinforcement. After behavior is well established, a long-lasting CS is presented. Unlike the CER situation, this CS does not signal the actual delivery of a shock, but rather signals the presence of a punishment contingency in which every response is accompanied by both food reward and a brief shock. This situation provides a clear marker for the punishment contingency, and shock levels and food motivation can be varied to maximize or minimize the level of conflict.

Two way avoidance

Finally, it should be noted that conditioned fear can interfere with fear motivated behavior as well as with positively reinforced behaviors. One situation in which this is especially salient is the two-way avoidance situation. In this task, a CS such as a light is presented in one end of an alley, followed by the delivery of foot shock. The subject can escape (or avoid) the shock by shuttling to the other end of the alley. After a period of time, the CS is presented in the other end of the alley, and the subject must return to the original location in order to escape or avoid the shock. Thus, there is no actual safe location, but rather the organism must learn that the CS signals the onset of "local" shock which can only be avoided by returning to yet another location in which shock already has been experienced. Of particular importance in considering this type of behavior is that increases in the amount of fear (i.e., higher shock intensities) actually slows down the rate of learning (cf., Moyer & Korn, 1964). The conflict in this situation interferes with learning to such an extent that typical rats require dozens or hundreds of responses to learn the task, while many do not learn at all.

The Human Condition

The experimental procedures described above, along with many variations, have been used extensively in basic research related to learning and the aversive control of behavior. Although it is an oversimplification, these tasks bear a reasonably close relationship to the various categories of psychiatrically important fears that may be encountered in the clinic.

The simple conditioning procedure can set the stage for both the normal, benign fears of everyday life and the more debilitating phobias. The distinction lies primarily in the time course and severity of the conditioned fear, as well as the object of the fear. In many cases, the conditioned fear response is only weakly established and transient, owing to the lack of a consistent relationship with a strongly aversive stimulus. Such fears are of little consequence. However, a strongly based fear of a common object or situations (e.g, elevators, bridges, cats, etc.) can be maintained indefinitely and even strengthened over time, owing to the individual's ability to avoid contact with the feared object.

In situations in which the object of the phobia cannot be avoided, the resulting influence on behavior is comparable to that observed in the CER procedure. The fear that results from the presence of the CS interferes with virtually all ongoing behaviors. This lack of behavior is not only debilitating in and of itself, but prevents the occurrence of behaviors that might normally lead to the extinction of the fear.

The clinical etiology of compulsive behavior is considerably more complex, but many cases may have their roots in simple Pavlovian conditioning. The disorder is complicated by the interaction of the learned fear response with overt behavior. Just as the punishment procedure described above is effective because the behavior itself comes to serve as a CS that signals an aversive consequence, behaviors that are a part of the compulsive repertoire can serve both to elicit the fear and then to reduce it, setting up a vicious cycle.

Vague or nonspecific anxieties are perhaps the most common form of debilitating fears. As the terms implies, there is frequently some degree of uncertainty about the actual source of fear. Furthermore, these vague fears can build upon themselves, such that individuals begin to fear that certain situations may lead to fear. In Pavlovian terms, this would be fear that anticipates the arrival of a CS that signals an aversive event. In human terms, it is the "fear of fear itself" that seems to be particularly dangerous.

C. THE BODY'S RESPONSE TO FEAR AND ANXIETY

The Adrenal Flight or Fight Response

The presence of a fear eliciting stimulus, whether it be a Pavlovian CS or an aversive stimulus per se, can produce a remarkable set of changes in the body's physiology. This set of changes is orchestrated primarily by the autonomic nervous system as outlined by Cannon (1929) many years ago. As a result of the diffuse action of the noradrenergic neurons of the sympathetic nervous system and the outpouring of catecholamines, (epinephrine, norepinephrine, and dopamine) and peptides from the adrenal medulla (see Fig. 4-3) the organism is prepared for intense physical activity. The heart rate and blood pressure are increased to provide more effective circulation of oxygen and energy sources. Blood is shunted away from the viscera to provide more circulation to the large muscle groups, and away from the skin surface to reduce bleeding that might occur as a result of injury. The blood chemistry changes to reduce the clotting time. Glycogen is released from the liver to produce an elevation of blood sugar and the spleen releases additional red blood cells into the general circulation to enhance oxygen carrying capacity. The respiration rate increases as does the effective volume of the lungs. The peptides (cf., Chap. 7 discussion of endorphins) lead to a direct reduction of pain. In general, there is a rapid mobilization of bodily resources to increase both the vigor and the intensity of the organism's response to the aversive situation.

The pattern of reactions described above can be elicited by a wide variety of situations, the major criterion being a situation that offers real or perceived danger. The physiological changes that result have clear, adaptive value by virtue of increasing the likelihood of successfully fleeing or fighting off the aversive situation. Indeed, any local folklore contains at least a few anecdotes of nearly superhuman feats that were accomplished under the influence of the sympathetic stress response.



General Adaptation Syndrome

The beauty of the adrenal stress response lies in the speed with which it prepares the organisms for action, but the resulting changes in physiology simply cannot be maintained for long periods of time. Hans Selye looked beyond this immediate response to stress and made two very important observations: (a) Long term exposure to stressful situations can deplete the organism's ability to maintain the stress response, and (b) The pattern of these deleterious effects is independent of the source of stress. Selye (cf., 1956) outlined a three-stage progression of responses to stress that he termed the General Adaptation Syndrome: Alarm, Resistance and Exhaustion. When a stressor is first encountered, a series of responses is initiated in the autonomic nervous system, the immune system and other defenses to cope with the emotional, behavioral and physiological aspects of the stressor. This is called the Stage of Alarm. The maintenance of this reaction to the stressor, which includes reparative processes such as fever regulation, tissue repair, control of inflammation, etc., is termed the Stage of Resistance. In some cases, the stressor cannot be successfully countered, and the organism enters the Stage of Exhaustion. In this stage, the defenses against the stressor begin to fail, metabolic reserves are depleted, there is a general decline in physiological functions, and serious illness or death ensues.

One of the most important of Selye's observations was that this is a general response that is independent of the situation that initiates it. The three stages of the General Adaptation Syndrome can be triggered by disease, injury, psychological stress, or some combination of these.

Surgical Shock

One of the common sources of trauma that can initiate the stress syndrome is that associated with surgical procedures. Even before the time of Selye, surgeons recognized the dual hazards of their art. Death can result either as a direct effect of surgical complications, or as a result of surgical shock that is not directly attributable to the success of the surgical procedure. The French surgeon, Henri Laborit, became interested in this phenomenon in the 1940's and undertook a program of clinical research and observation that was to have far reaching consequences for the treatment of stress related disorders (cf., Caldwell, 1970).

Laborit recognized that surgical trauma involved intense activation of the autonomic nervous system. Normally, the autonomic nervous system maintains bodily functions within fairly tight limits, automatically adjusting the organism's physiological needs to fit the ongoing requirements. These routine adjustments are primarily the responsibility of the parasympathetic, or vegetative, division of the autonomic nervous system (see Fig. 4-4). But in times of severe stress, these systems can run amok, producing bodily changes that are counterproductive, leading to the life threatening condition that is commonly referred to as shock. Attempts to treat the stress may, in some cases, contribute further to the stress. Laborit stated this with an eloquence that survives translation:

"In fact, perfect lytics are not yet at our disposal and even if one existed, it probably would be effective only in large doses. In that case, an injection of the drug would increase the stress that, when it attains a certain level, elicits organic defense reactions that are quite contrary to our fixed goals (prevention or mitigation of those exaggerated reactions that defend our invariant inner milieu that guarantees liberty but not always life.)

(trans. by Caldwell, 1970, p. 29)

Sudden Death



Laborit was not alone in challenging Cannon's sympathetic model of stress. In a paper that was originally published in 1942, Cannon had suggested that massive overreaction of the adrenal system could lead to Voodoo death (sudden death that was caused by emotional rather than physical stress). The most impressive evidence against this model came from an elegant series of experiments performed by a psychologist, Curt Richter, who investigated this curious phenomenon of sudden death.

Richter's initial experiments bore little or no relationship to the stress syndrome. He had become concerned that the methodical inbreeding of the albino laboratory rat had rendered it too weak to serve as an adequate model subject. He attempted to prove his hypothesis by showing that the albino rat was physically weak when compared to its wild, Norway rat counterpart. He developed an endurance test that involved swimming in a circular tank, equipped with a sort of whirlpool in the center that ensured continuous swimming. The results of the first experiment were somewhat curious: At optimal water temperatures, most of the rats swam 60-80 hours, but a few died within 5-10 minutes. Why? Richter recalled an earlier observation in which a rat's whiskers (vibrissae) had been trimmed as part of another experiment. The rat began to behave strangely, and died about eight hours later! Now, Richter suspected that this might have been related to stress, and clipped the whiskers of 12 rats before doing the swim test. Three of the 12 died within minutes, but the remaining nine swam 40-60 hours. By contrast, all wild rats tested in the same way died within minutes and many of them die even without the whisker clipping!

Richter searched beyond the superficial aspects of these results. He suspected that this sudden exhaustion and death of the wild rats might be related to the Voodoo death phenomenon, as suggested by Cannon. The prediction to be made by Cannon's sympathetic model was clear--the release of adrenaline should cause the heart to beat faster and faster until it no longer had time to fill between beats, leading to death in systole (i.e, a contracted heart). The actual results were exactly opposite. The heart rate of the wild rats became slower and slower, with the autopsy showing the heart to be completely engorged with blood. These results bore all the earmarks of a massive parasympathetic response.

Richter tested the notion that this was a parasympathetic response using two pharmacological procedures. In one case, he administered mecholyl (a parasympathetic mimicker) to the albino rats. They quickly acquiesced to the swimming task and sank to the bottom, like the wild rats. In the other case, he administered atropine (a parasympathetic blocker) to the wild rats, which prevented the sudden death in some, but not all of the rats tested. The combination of these results, summarized in figure 4-5 spun an irrefutable conclusion: The sudden death phenomenon was parasympathetic. Why?

Richter pursued the emotional causes of this stress syndrome. Is it possible that the normal, sympathetic response to stress is replaced by a parasympathetic response under extreme conditions? The rats' vibrissae provide a major source of information. Lacking this information in a hostile environment such as the swimming tank, could render the situation hopeless, leading to this paradoxical parasympathetic response. But what about the wild rats? Richter suggested that they may also view the situation as hopeless simply because (being wild) it is more stressful to be handled, and they have never before been in captivity. To test this notion, he allowed several of the wild rats to sink to the bottom of the tank. Then, retrieving them from otherwise certain drowning, he placed them on the table until they recovered, then put them back in the tank. After a few repetitions of this lifeguard routine, the wild rats would swim for many hours. The conclusion, which seems valid, was that the wild rats learned that the situation was not hopeless after all.

The results of Richter's experiments bring up several important points that go beyond the analysis of the stress syndrome:

1. A behavioral phenomenon can be blocked through the pharmacological blockade of the target organ receptor. (Atropine prevented the sudden death in wild rats).

2. A behavioral phenomenon can be mimicked or exaggerated through the pharmacological stimulation of the target organ receptor. (Mecholyl triggered the sudden death in albino rats).

3. Manipulations that change the perception of the environment can either exaggerate a behavioral phenomenon (as in the case of shaving the rats' vibrissae) or block a behavioral phenomenon (as in the case of rescuing the wild rats).

4. The perception of the environment is an important determinant of the nature of autonomic response to stressors.

Ulcers

Executive monkeys


The hallmark of stress disorders is the formation of ulcers. This condition has become synonymous with demanding job situations such as executive positions, and with other situations that involve daily exposure to stressful conditions. The superficial reason for ulcer formation is the release of stomach acids into an empty stomach. The presence of these digestive juices, along with some local vascular changes, lead to the digestion of the stomach lining itself, and can sometimes lead to an actual hole through the stomach wall, a perforated ulcer. The real reasons for ulcer formation, however, can be traced back to the emotional responses that set the stage for this untimely release of digestive juices.

Ulcers are far more than a clinical curiosity. They are painful and even life threatening to the individuals who are afflicted. Furthermore, they account for tremendous financial losses in terms of workdays lost and medical costs. The impact of this disorder has stimulated a great deal of research to determine the cause of the disorder and to develop pharmaceutical treatments for the disorder. Obviously, the best solution would be to eliminate the conditions that initiate the ulcerative process, and toward this end, there has been considerable effort to develop an animal model of the stressful conditions that cause hypersecretion of gastric acids.

The cornerstone of this effort was Brady's (1962) so-called Executive Monkey study. This study is important for historical reasons, even though the basic conclusions drawn from the study were, ultimately, shown to be exactly opposite to current knowledge in the area. Brady trained a group of monkeys to perform a free operant (Sidman) avoidance task which required that a lever be pressed to avoid shock to the tail. If the monkeys allowed too much time to elapse before pressing the lever, an electrical shock was delivered to the tail. The executive monkeys spent each workday sitting in the restraining chair performing this task. The worker monkeys sat in a similar restraining chair with electrodes attached to their tails, but the delivery of electrical shock was entirely dependent upon the executives' decisions. If the executive received a shock, so did the worker. Consistent with the predictions, the executive monkeys eventually developed gastric ulcers and the worker monkeys did not. Unfortunately, these results support the wrong conclusions because of a combination of procedural details and flaws in experimental procedure. We will return to an analysis of these results later.

The triad design

The most comprehensive behavioral research in this area has been done by Weiss and his associates (e.g., 1968; 1981). These experiments, utilizing rats as subjects, have reached conclusions that are diametrically opposed to those of Brady, but at the same time confirm the actual results of those early studies. Although the testing procedures have varied over the years, most of these experiments have utilized the apparatus and procedures that are shown in Figure 4.6 and outlined below.

The rats were placed into small restraining cages with electrodes attached to their tails. A small wheel, located immediately in front of the rats, could be turned with their front feet. In the prototype situation, there were three testing conditions (a triad) that differed in terms of the degree of interaction each rat had with the shock. Although each of these experiments involved many rats, the testing was always conducted in triads so that the environmental conditions of the rats were interdependent. The experiments will be described separately to demonstrate the major conclusions that were reached by Weiss' research group.

Control of stressors

The most critical set of experiments involved an assessment of the importance of control over the environment. The test triad in these experiments was exposed to the following conditions:

(a) Escapable: Electric shock was delivered to the rat's tail at random intervals. Once the shock was begun, it was programmed to continue until the rat turned the wheel with its front paws. Thus, the rat had control over the termination of the shock.

(b) Inescapable: This rat's tail was connected to the same shock source as the experimental rat. Although it could turn the wheel, the wheel did not influence the shock. Shock termination only occurred when the experimental animal successfully turned it off. This link to the behavior of another subject is referred to as a yoked control procedure.

(c) Control: This rat was maintained in the restraining cage for the duration of the experiment, but was not exposed to the electric shock.

The results of these experiments were clear: The rats in the yoked control condition developed gastric ulcers, the other two groups did not. Contrary to the results of Brady's experiments, the subjects that were in charge of shock decisions were the ones that developed the ulcers. However, if these results are described in slightly different language, they seem to make a lot more sense. The rats that had control (i.e., mastery) over the shock were less stressed than those which were at the mercy of their environment.

Prediction of stressors

A second set of experiments extended Weiss' analysis of the conditions that lead to ulcers. In these experiments the following conditions formed the test triad:

(a) Signaled: A signal (CS) was presented at random intervals, followed by a brief, inescapable shock.

(b) Unsignaled: Again, the rats in this condition received shocks that were identical to those received by the experimental group. The distinguishing feature was that they did not receive the CS that signaled the impending shock.

(c) Restrained Only: These rats received neither the CS nor the shock.

The results of these experiments began to support a more general notion of mastery over the environment. Once again, it was the subjects in the yoked control condition that developed the severe ulcers. The experimental animals received the same shock, but apparently the mere knowledge of when the shock was going to be delivered reduced the stress. They developed very few ulcers.

Presence of Conflict

A third set of conditions begins to come closer to the human conditions that are likely to engender ulcer formation. This is the presence of conflict, which Weiss modeled with the following triad:

(a) Signaled escape: A signal (CS) for impending shock was presented at random intervals, as in the experiments investigating the importance of prediction. However, these rats also had control over the shock, in that turning of the wheel could either terminate the shock or, if it occurred during the CS, actually avoid the shock altogether.

(b) Conflict: These rats were exposed to conditions that were identical to those of the experimental rats, except that on some trials, the wheel turning response itself was punished with electric shock.

(c) Restrained Only: Again, these rats were simply restrained for the duration of the experiment.

The rats in the signaled escape condition of this experiment were completely free from ulcers. As suggested by the separate experiments above, the presence of both prediction and control negates the formation of ulcers. The presence of conflict, however, led to severe ulceration. In some sense, it would appear to be better to have no control or prediction at all, than to have these available but inconsistent. Figure 4-7 summarizes these results.

Stressors Revisited

The results of these experiments support a remarkable conclusion: Noxious stimuli are not inherently stressful. In all of the experiments above, the experimental group received shock that was identical to that of the second group in terms of the interval of presentation, the intensity and the duration. The critical factor was not the presence or absence of electric shock, but rather the presence or absence of what we might call a particular "interpretation" of the electric shock. Prediction, control, and the absence of conflict are the three factors that prevent noxious stimuli from becoming stressors.

Why did Brady get the opposite results? The answer lies within Weiss' experiments. Animals that are exposed to shock (even though it is neither predictable nor controllable) will not develop ulcers unless the frequency of occurrence is fairly high--an occasional brief shock is simply not stressful enough to cause a problem. In Brady's experiments, the executive monkeys were skilled enough to prevent most shock from occurring, so the worker monkeys were not exposed to very many shocks. There are several reasons for the development of ulcers in the executive monkeys. Even though they had control, the free operant situation requires constant vigilance, and there is no external CS to predict the shock. The sessions lasted for hours and the constant requirement of timing responses to avoid shock is obviously stressful. Another important factor was that all the monkeys were initially trained in the executive condition, and when about half of the subjects had mastered the task, the remaining subjects were switched to the worker condition. This biased selection of subjects made it even more likely that the executive group would develop ulcers, because later studies with rats have shown (for reasons that are not clear) that rats which learn avoidance responses quickly are also more prone to develop ulcers.

All of this is consistent with the conditions that lead to ulcers in the human environment. The prediction and control of corporate executives is illusory. Although they are required to make decisions, the environment is sufficiently complex that the outcome of the decisions is uncertain and occasionally punished (hence, conflict). It is the menial laborer who has prediction and control by virtue of simple tasks, scheduled daily activities, and known outcomes for most work related behavior. Not that these individuals are immune to ulcers, but the source of the conditions that lead to the ulcers is more likely to be found in the home or social environments of these individuals than in their work places.

The experimental procedures that result in ulcer formation fit into a larger context of situations that produce aberrant responding of the autonomic nervous system. The procedures that produce ulcers do not appear, on the surface, to be life threatening. When compared to the trauma of either a surgical procedure or Richter's swimming task, the lack of prediction or control over electric shock would seem to be rather benign. Yet, the common emotional fabric of all of these is the hopelessness and lack of control of the environment. It is the behavioral interpretation of the environment (be it valid or not) that leads to an autonomic imbalance in the direction of parasympathetic over-responding.

D. THE PHARMACOLOGY OF STRESS RESPONSES

The Search for Autonomic Stabilizers

We return now to the Val-de-Grace Hospital in France, where Laborit had been continuing his work on the pharmacological control of surgical shock. Laborit and other surgeons had already used atropine (another compound that blocks acetylcholine at the target organ) to improve the recovery from surgical shock. This treatment was only marginally effective, so Laborit began using low dosages of curare as a ganglionic blocker. As shown in Figure 4-8, curare has little or no effect on the autonomic target organ receptors, but effectively blocks the nicotinic receptor for acetylcholine that is present in both the sympathetic and parasympathetic ganglia. Thus, curare tended to decrease both the sympathetic and parasympathetic reactivity of the autonomic nervous system. This procedure (sometimes used along with atropine) was somewhat effective, but many patients still went into surgical shock.

One of the effects of surgery (or other tissue damage) is the release of histamine, which is also a potent stimulator of some autonomic target sites. (A common example is the redness of the skin that occurs through local vascular responses when it is scratched.) This response to tissue damage (which Laborit called "silent pain") occurs under anesthesia as well as when an individual is awake, adding to the complications of surgery. By the late 1940's, several antihistamine compounds had been developed, and were being used with some degree of success to control surgical shock. Laborit was searching for what he termed the perfect lytic compound--a drug that would stabilize the autonomic nervous system and, in a sense, "dissolve" the patients' fears. He was somewhat pessimistic, however, because he recognized that a heavy dosage of a drug is itself a stressor that can trigger the stress syndrome (see his quote above).

Despite his pessimism, Laborit saw hope for a lytic compound in one of the antihistamines, namely promethazine. In addition to its effects of stabilizing the peripheral autonomic nervous system, the drug also had mild effects on the central nervous system, resulting in a sort of indifference to the stressful environment. This indifference was in contrast to a troublesome sedative and hypnotic side effect that accompanied many of the other antihistamine compounds. Caldwell (1970) relates an instance in which one of Laborit's patients ran through a red light, even though he was not noticeably drowsy and inattentive. Working with a biochemist in a drug company (Specia), Laborit guided the manipulation of antihistamine molecules to bring about maximal central activity, irrespective of action in the periphery. Finally, on December 11, 1950 the drug that was to launch modern psychopharmacology was synthesized: That drug was chlorpromazine.

The Tranquilizers (Phenothiazines)

Chlorpromazine was not, perhaps, the perfect lytic that Laborit was seeking, but it was close. When administered to patients prior to surgery, the effect was remarkable. The drug did not cause heavy sedation, thus allowing the patients to remain aware of their environment. They could carry on conversations, answer the physician's questions, and clearly were in contact with their environment. But the drug did cause a certain indifference to stressful stimuli, greatly reducing the normal preoperative fears, reducing the amount of anesthesia that was necessary to conduct the surgery, and most importantly, reducing dramatically the likelihood of death resulting from surgical shock. Virtually all of these effects are caused through action on the brain rather than the peripheral nervous system. It was an autonomic stabilizer that worked by virtue of changing the perception of the environment. It was, in the words of Laborit, a Pavlovian deconditioner--stimuli that previously elicited fear were as benign following chlorpromazine administration as they would be if experimental extinction had taken place.

Literally thousands of experiments have been done to test the effectiveness of chlorpromazine, but the acid test in terms of animal experiments would be Richter's swimming test. If the drug is truly effective as an autonomic stabilizer, then it should prevent the sudden, parasympathetic death of rats in the swim test: It did.

The first patient to be treated with chlorpromazine was a young man who had a history of agitated, psychotic behavior. He had entered the Val-de-Grace Hospital in September of 1949 and received 15 shock treatments. In February of 1951, he returned to the hospital and received 24 additional shock treatments (both insulin and electric). In January of 1952, he was given 50 mg of chlorpromazine and immediately became calm. After seven hours, his agitation returned, but subsided again with a second dosage of the drug. Gradually, the drug's effectiveness lasted longer and longer, and the patient was released after 20 days.

It is almost impossible to overestimate the impact of this drug and the related phenothiazines on the care and treatment of psychiatric patients. Prior to the advent of chlorpromazine, psychiatric patients were rarely released from the hospital. The chronic, in-patient population was ballooning, and the care bordered on the barbaric. Straight jackets and restraining chairs were used routinely for the protection of patients and staff alike. Electric and insulin shock treatments were common procedure. There were no alternatives and the patients were more likely to get worse than to get better. Chlorpromazine literally freed the psychiatric patients from their bondage. It effectively reduced their fears and agitation to the point that restraining devices were unnecessary. The drug was not habit forming and tolerance was minimal. Most importantly, the patients were not asleep as they had been with barbiturates and other sedative/hypnotics. They retained their ability to interact with their environment, but were indifferent to the stressors.

With the advent of chlorpromazine, patients went home. As shown in Figure 4-9, their is a dramatic reversal of the in-patient population beginning in 1952. The savings in dollars has been estimated in the billions, and the savings in human suffering is incalculable. The patients were not, to be accurate, cured. But the drug allowed them to regain a sufficiently cogent interaction with their environment to be taken care of safely in a family setting.

The details of the action of the phenothiazines will be presented more fully in Chapter 6, but it is important to consider the development of the drugs at this point because of the impact they had on the investigation of the pharmacology of stress. The immediate success of chlorpromazine made drug therapy in psychiatry a reality, and spawned a major search within the pharmaceutical industry for more, if not better, compounds. As a result, chlorpromazine is simply the prototypical example of a group of chemicals known as phenothiazines, which are sometimes referred to as neuroleptics (in reference to their autonomic stabilizing effects), as major tranquilizers (in reference to the Pavlovian deconditioning effects), and as antipsychotics (encompassing both of the above and the fact that they are especially effective in treating this patient population).

The Antianxiety Drugs (Benzodiazepines)

The success of chlorpromazine in treating psychotic patients led to an intense search for other drugs that would have a calming influence, particularly on the fears and anxieties that occasionally interfere with the lives of otherwise normal individuals. The phenothiazines were, to some extent, too much of a good thing. The emotional flattening and autonomic side effects were reasonable alternatives to psychotic episodes, but seemed like a high price to pay for the treatment of patients who were, perhaps, a little nervous about their new job. Consequently, the search for new drugs was aimed toward compounds that would calm the day-to-day anxieties while having only minor side effects. The most successful drugs produced by this effort was a class of compounds known as the benzodiazepines, of which chlordiazepoxide (Librium) and diazepam (Valium) are the most commonly prescribed.

These compounds are variously referred to by the name of the chemical class, as minor tranquilizers, and as antianxiety compounds. They are useful and widely prescribed to reduce the tensions and anxieties associated with job and family situations, as well as to relieve or prevent associated problems such as muscle tension and headaches.

The screening of drugs that are potentially useful in treating stress related disorders virtually requires animal models. The financial costs, time requirements, and potential dangers of clinical tests with humans all require that the initial stages of testing be done with animal tests. As a result, there are several testing procedures that are useful in categorizing the drugs and to provide further information about the nature of the behavioral changes produced by the drugs' actions on the brain.

In the discussion above, it was pointed out that one of the major effects of chlorpromazine was something termed Pavlovian deconditioning. The results of animal tests confirm this notion, and it is worthwhile to directly compare the effects the phenothiazines and the benzodiazepines on these types of tests. In appropriate dosages, chlorpromazine (and other phenothiazines) can reduce avoidance responding (i.e., conditioned responses to fear), while leaving escape behavior intact (cf., Cook & Sepinwall, 1976; see Fig. 4-10A).

This selective effect on these two closely related responses provides an excellent initial screen for drugs that are likely to share the antipsychotic effects of chlorpromazine in the clinic. By contrast, the anti-anxiety compounds reduce avoidance behavior only in dosages that are sufficiently large to also impair escape responding (Fig. 4-10B). This nonspecific effect can be obtained by several different classes of drugs (e.g., those that simply impair movement), so this task has little or no utility in screening for new compounds that might serve as anti-anxiety drugs.

There is, however, a task that provides a sensitive screen for potential antianxiety drugs. These drugs seem to be uniquely effective in changing performance in the Geller-Seifter punishment procedure that was described earlier. Initially, this test was used to demonstrate the specific effects of barbiturate drugs, because these were the most widely prescribed drugs for the treatment of anxiety. The specific response to punished responding following barbiturate administration is mirrored by the administration of chlordiazepoxide and other benzodiazepines In this test, the animals that have been treated with the drug show perfectly normal behavior patterns in the food rewarded portion of the schedule, but are markedly different from control animals during the punishment portion. Whereas normal rats will stop responding when the signal for shock plus food is presented, rats that have been treated with one of the anti-anxiety compounds are released from this suppressive effect and continue their high rate of responding.

The Geller-Seifter screening procedure is especially important because it discriminates the anti-anxiety compounds from other classes of drugs. Chlorpromazine and other antipsychotic compounds are ineffective in this procedure. General depressants (e.g., barbiturates) or stimulants (e.g., amphetamine or caffeine) of the central nervous system may alter the punished responding, but only in dosages that have a comparable influence on the food rewarded portion of the schedule.

The Geller-Seifter procedure is not the only method for screening drugs for their antianxiety properties. In fact, this method is so cumbersome and time consuming that its use tends to be limited to those situations that require an especially rigorous test of a drug. Other tests which are, perhaps, not so sensitive are much easier to use. For example, chlordiazepoxide will increase the amount of novel food that a rat will consume (Poschel, 1960; in Sepinwall & Cook, 1978). This is apparently not related to any changes in hunger per se, but rather to the more general response to novel (mildly aversive?) situations. When a rat is exposed to a novel environment, there is an increase in plasma corticoid levels. This index of stress can be effectively blocked with administration of minor tranquilizers (e.g, Lahti & Barshun, 1974).

Finally, a particularly easy method of measuring the response to punishment has been shown with the consumption of salt solutions. Although rats show a positive taste response to a hypertonic solution of sodium chloride, the drinking of this solution is rather quickly limited by the aversive postingestional consequences (the animal becomes thirsty as a result of drinking). The administration of minor tranquilizers will increase the amount of hypertonic salt solution that is consumed (e.g., Falk & Burnidge, 1970).

From Laboratory to Clinic and Back

Receptors for Phenothiazines


The pathway from the biochemist's laboratory to the clinician's administration of a drug is not a one way street. Although some of the screening tests may have face validity, there is always a danger that the aspect of the drug that causes an effect on a screening test is not always the same as the one that causes its clinical effectiveness. This problem can never be eliminated completely, but the level of confidence can be raised when tight relationships emerge on the basis of extensive use of the drug in humans. If a drug or class of drugs has been used extensively in the clinic, it may be possible to make a direct comparison between the clinical results and some laboratory screening procedure. In the case of the tranquilizing drugs, there are two such relationships that are especially instructive.

As shown in Figure 4-11, there is a very strong relationship between the clinical dosage of the various antipsychotic compounds and the ability of these compounds to replace another molecule (haloperidol) from dopamine receptors. The logic is as follows: If the dopamine receptors of a test object are simultaneously exposed to haloperidol and some other compound, the two drugs will compete for the receptor sites. For the sake of illustration, if 100 molecules of a compound that has a strong affinity for the dopamine receptor is pitted against haloperidol, perhaps as many as 80 of these molecules will be successful in occupying dopamine receptor sites. If a weak compound is used against haloperidol, then perhaps only 20 molecules would be successful. In order to get 80 molecules of the weak compound into the receptor sites, a higher dosage (in this example, 400 molecules) would have to be used. (It should be noted that this test is based on the D1 receptor for dopamine; see Chapter 8 for further discussion of D1 and D2 receptors.)

In the clinic, the mechanism of action of the drug may not be known, and efficacy is based upon the relief of symptoms. Drugs that are weaker must be prescribed in larger amounts than drugs that are stronger. When a class of compounds has been given to thousands of patients and dosages have been adjusted, then the compounds can be ranked in terms of their relative strength, or potency. Note that this does not necessarily mean that any one drug is better than another, but simply that some drugs are more potent than others (the same relationship would hold if a single drug were "watered down" so that a larger amount would have to be given to achieve an effective dosage). The observation of interest is depicted in figure 4-11 (after Creese et al, 1976). When the phenothiazines are rank ordered in terms of their clinical potency, the list is virtually identical to that obtained when they are rank ordered in terms of their affinity to the dopamine receptor. In other words, the potency of a drug to bind to the dopamine receptor is closely related to the potency of that drug to relieve psychotic symptoms in the clinic. It takes a hard nosed skeptic to believe that this would occur by chance.

In the case of the minor tranquilizers, a comparable relationship can be shown between the clinical potency of these compounds and their effectiveness in blocking the suppression of punished responses in the Geller-Seifter procedure. As schematized in figure 4-12, drugs that must be given in large quantities to produce the desired clinical effect must also be given in large quantities to change the behavior in the Geller-Seifter procedure.

Receptors for Benzodiazepines

The relationship of the benzodiazepines to neurotransmitter systems remained elusive for many years. These drugs do not significantly alter the brain concentrations of dopamine, norepinephrine or serotonin, although the turnover rate of all of these is reduced. Over the years, the compound known as GABA (pronounced gabbuh; short for gamma amino butyric acid) has gained increasing respect as a neurotransmitter. It is present in virtually every portion of the brain, it has consistently inhibitory effects by virtue of opening chloride channels (cf., Chapter 8), and it is probably the single most plentiful neurotransmitter in the brain (see Olsen, 1987 for discussion). The receptor for GABA has been termed the GABA receptor complex (see Fig. 4-13) and is one of the most interesting developments in neurochemistry. It would appear that there are three interacting receptors on this site: One of them is the primary GABA receptor, which regulates the Cl- channel. The second is a receptor that responds to sedative and convulsant drugs. The third is receptive to benzodiazepines, and their presence enhances the normal activity of GABA.

There is some possibility that the brain produces endogenous compounds that are comparable to the benzodiazepines. The evidence for these naturally occurring substances is threefold: (a) labeled diazepam is tightly bound to specific receptors, (b) the rank order of clinical potencies of the benzodiazepines is highly correlated with the rank order of the ability of these compounds to displace the labeled diazepine from these receptors (see Fig. 4-14; after Baestrup & Squires, 1978), and (c) exposure to stress appears to block the binding of benzodiazepines, presumably because the sites already have become occupied by some stress induced substance. Furthermore, the rank order of clinical potencies is the same as the rank order of the ability of the compounds to displace this labeled compound from the receptor (cf., Moehler & Okada, 1977; see Fig. 4-15; after Lippa et al, 1978).

Anticholinergics as Anti-punishment Drugs

Drugs that block the effects of acetylcholine on parasympathetic target organs are also extremely effective in blocking the effects of punishment, nonreward, and the debilitating effects of conflict--in animals. Although both scopolamine and atropine have been used in some clinical situations (especially presurgically,) they have not been as useful clinically as the animal data would suggest.

There is a great deal of evidence that the brains systems that are involved with reacting to punishment and nonreward utilize acetylcholine as the neurotransmitter (cf., chapter 2 and Carlton, 1963, for related discussion.) Both atropine and scopolamine (cholinergic blocking agents) alter the behavior of rats in a variety of related situations: Behavior that is punished with shock persists. Behavior that is no longer reinforced persists. Stimuli that signal a temporary period of nonreinforcement (time-out experiments) are ignored. Schedules that require low rates of responding to obtain reward (drl schedules) cannot be mastered. These results have been observed in different laboratories, using different reinforcers and other testing parameters, and in different species. The conclusion that cholinergic blocking agents reduce the response to punishment and nonreward is almost inescapable.

These compounds have also been used in two other situations that seem even more relevant to the reduction of stress responses. One of these has already been discussed: Atropine injections blocked the sudden death phenomenon in Richter's swimming task. The other involves the two way avoidance procedure. Normal rats have great difficulty learning this task, presumably because a successful avoidance response requires that the rat return to a location in which shock (or a signal for shock) has just been experienced. Scopolamine or atropine dramatically increase the ability to master this task, presumably because it reduces the disabling response to conflict.

When this discussion was begun, it was asserted that these drugs influence the brain systems that control the responses to punishment and nonreward. The evidence for this assertion is strong, but at the same time provides a clue concerning the limited usage of these drugs in the clinic as stress inhibitors. Perhaps the major reason why atropine and scopolamine are not suitable for routine administration in humans is because they are too effective in the periphery. Recall that Laborit had used scopolamine as a presurgical treatment prior to the development of chlorpromazine, it was effective in blocking the strong parasympathetic component of surgical shock. Likewise, this type of cholinergic blockade was effective in blocking the sudden death phenomenon in Richter's studies. However, the potency of these compounds in blocking the parasympathetic effector organs is itself a liability. In the case of diminishing surgical shock or preventing voodoo death, certain undesirable side effects can be tolerated. But for routine administration, the accompanying dry mouth, dilated pupils, decreased gastrointestinal activity and other autonomic effects are undesirable.

The progression of drugs that were used in the prevention of surgical shock provides a particularly good lesson of pharmacological principles. Scopolamine and atropine block the effects of acetylcholine at the receptors of the actual target organs (i.e., the smooth muscles and glands) of the parasympathetic system. In other words, the "command system" of the autonomic nervous system may remain functional while the final response is blocked. Laborit went back one step and blocked the action at the autonomic ganglia with low dosages of curare. This resulted in an autonomic stabilizing effect, by reducing activity of both the sympathetic and parasympathetic divisions of the autonomic nervous system. This type of action also had its limitations, because it was, in some sense, masking the final stages of a stress reaction that had already been initiated in the central nervous system. Laborit was seeking a drug effect that would block the initial stress interpretation in the brain, and found this effect in chlorpromazine. The major point here is that it is preferable to forestall the stress reaction in its initial stages than to allow it to develop and then block its effect at some later point along its synaptic route.

The problem with scopolamine and atropine is not that they lack central effects, but that they have both central and peripheral blocking activities. In fact, there is strong evidence that the major influence of these compounds on the tasks outlined above is attributable to their effects upon the brain rather than the autonomic effectors. Both of these compounds are amines and in their normal states the nitrogen on the side chain has three radicals attached to it and is neutral. These compounds can be transformed biochemically by adding a fourth radical (methyl) to the nitrogen leaving it with a positive charge. The resulting compounds (called quaternary amines) are commonly referred to as methyl atropine and methyl scopolamine and have the very useful property of being virtually unable to penetrate the blood brain barrier (cf., Chapter 3.) This property means that nearly all of their blocking effects are restricted to the peripheral parasympathetic effectors, while brain acetylcholine systems are left to function normally (see Figure 4-16).

A typical experimental design compares the behavior of a control group (saline injected) with that of a group injected with standard atropine and that of a group receiving methyl atropine. In virtually every experiment that has been done, the results are clear cut: Standard atropine reduces the response to punishment, nonreward and conflict, whereas methyl atropine (which has the same or even more potent peripheral effects) has no effect on these behaviors. What this means is that the blockade of the parasympathetic organs plays little or no role in the effects of these drugs on stress related behaviors. Virtually all of the effects can be attributed to their action on the brain. In this regard, it would be very interesting to know if methyl atropine would prevent the sudden, parasympathetic death in Richter's swimming task (it probably would not) and if a form of scopolamine that worked only on the brain, but not the periphery, would be a useful drug in the treatment of clinical stress disorders (it probably would). In any event, we are not yet finished with the role of the autonomic nervous system in stress responses, and we soon will see evidence that the peripheral responses are considerably more important than they were once thought to be.

Treatment of Ulcers

As in the more acute instances of shock reactions, the formation of ulcers can be blocked or retarded by the injection of cholinergic blocking drugs such as atropine. It is not, however, the treatment of choice in the clinic for the same reasons as discussed above, namely, side effects. It is one thing to demonstrate the effectiveness of atropine by blocking the formation of ulcers in an animal experiment that lasts a few hours or a few days. It is quite another to use such a broad spectrum drug over a period of years in a human patient.

There are two pharmacological solutions to this problem that reflect importantly different therapeutic strategies. One of these, which we have seen above, is to counter the stress response at the developmental stages in the brain. In this regard, the antianxiety compounds are successful in both experimental models and in the clinic. Chlorpromazine might also be effective, but because of its potency is not routinely used for this purpose. Obviously, another even more desirable (and effective) approach is to eliminate the environmental conditions in the patient's life that lead to the formation of ulcers, but it is not always easy for the therapist to extricate people from their yoked control situations.

The second pharmacological approach is to basically ignore the stressful situation per se and very specifically block the final stage of the stress response at the gastric receptors. As discussed earlier, cholinergic blockade is not sufficiently specific, but there is an alternative. Once again, the roots of this alternative go back to Laborit's work on surgical shock. He referred to the silent pain of the surgeon's knife, recognizing that the tissue damage resulted in a large autonomic response. This was due to the stimulating properties of histamine (literally meaning amine from the tissues) on autonomic effectors. Recall that Laborit's search for an autonomic stabilizer centered on antihistamines, but most of these compounds had broad actions in both the central and the peripheral nervous systems. Over the years, the research that was spawned by these early problems led to the discovery of at least two types of histamine receptors, called H1 and H2 (see Douglas, 1980, for discussion). Of these, the H1 receptors are far more common, being involved in response to injury, hypersensitivity reactions (allergies), and other conditions. The H2 receptors are far less common, being primarily involved with the regulation of the volume and acidity of gastric secretion (see Figure 4-17). Thus, it is possible to administer an H2 blocking compound that will block the hypersecretion of ulcer producing stomach acids, while leaving most of the remaining activities of histamine unaltered. One of these compounds, cimetidine (trade name, Tagamet), has become one of the most widely prescribed drugs in the world!

E. THE AUTONOMIC RESPONSE: Chicken or Egg?

James-Lange Theory of Emotion

The autonomic nervous system has been linked integrally to emotions since Walter Cannon's classic description of the adrenal response. According to Cannon, and later formulations developed with Bard, the autonomic nervous system served rather like a support system. The organism recognized some emotion provoking stimulus in the environment, analyzed it, experienced the appropriate emotion, and as a result of these processes, triggered the autonomic nervous system into action for the ensuing flight or fight.

William James (1890) proposed an alternative view which, on the surface, seems totally unreasonable. The James-Lange formulation proposed that the emotion provoking stimulus triggered the autonomic nervous system directly (although there was a provision for central nervous system involvement), but the actual experience of the emotion lagged behind and depended upon a "reading" of the autonomic reaction. Popular (and overly simplistic) metaphors of this theory proclaim that an individual "...is fearful because he is running from a bear", or "...is angry because she hit somebody." This notion seems to have confused cause and effect.

Cannon (1927) pointed out a series of problems with James' view of the emotional experience: (a) The visceral response is slow to develop. (b) The viscera themselves are rather insensitive, even to physical trauma such as cutting or cauterization. (c) The same response (e.g., an elevated heart rate) can be elicited by fear, running around the block, or falling in love. (d) Patients with spinal injuries that lead to paralysis and loss of bodily sensations experience full emotions. (e) Injections of adrenaline do not result in emotional experiences. At the time of this argument, Cannon was perhaps the ranking physiologist of the world, and William James was merely a gifted writer, philosopher, and psychologist who was treading on the foreign soil of physiology. Cannon's view prevailed.

Cannon's professional stature overshadowed some of the weaknesses of his objections. However, the weaknesses became more and more apparent as additional information about the autonomic nervous system unfolded through the years. It is true that visceral changes are sluggish and slow to develop, but so is it true that the full emotional experience is often slow to develop. An all-too-frequent experience is the near miss of an automobile accident which almost instantly mobilizes complicated motor responses, while the full range and impact of the emotions may come seconds, minutes, or even hours later. It is also true that the viscera can be cut, cauterized and otherwise insulted during surgery with little or no sensation to the patient, but this is a moot point. We certainly can experience the rapid heartbeat, flushed skin, and butterflies in the stomach during emotional experiences. Cannon's point about the origins of an increased heart rate was also weak, in that he failed to recognize the possibility that different emotions engender different patterns of autonomic responses (cf., Ax, 1953; Ekman, et al, 1983; Funkenstein, 1955). Patients who lack the ability to move or to feel somesthetic stimulation of the body still retain a large portion of autonomic sensitivity via the cranial nerves, especially the vagus nerve. These patients also report a lack of emotional intensity, feeling "as if" they were angry. Finally, the experiments involving the effects of adrenaline injections were incomplete in design, missing an important point that even James missed. These studies form the basis for the remainder of this section.

Schachter and Singer's Model

The experiments of Schachter, Singer and their colleagues (e.g., Schachter, 1971; Schachter & Singer, 1962) have shed new light on the James-Lange theory of emotions. Their results show clearly that autonomic arousal can set the stage for (rather than being the result of) emotional experience, and elucidate some of the difficulties that other experimenters (including Cannon and James) have had in triggering emotional reactions with adrenalin injections. We turn now to a consideration of some of their results.

A typical experimental procedure employed by Schachter and Singer involves the injections of either adrenaline or saline (a placebo) and the presence or absence of an emotion provoking situation. In each study, the subjects who had received the injections were divided into two groups. One group simply filled out a questionnaire that contained some rather pointed items. The second group filled out the same questionnaire, but a confederate who pretended to be a subject vividly expressed his outrage at the nature of the questions, tore up the response sheet, and stomped out of the room. Post-test interviews showed the following pattern of results: (a) The questionnaire per se did not elicit anger for either the subjects injected with the placebo or those injected with adrenalin. (b) The subjects injected with the placebo did not experience anger, even when exposed to the confederate. (c) The subjects who had received adrenalin injections, however, were strongly influenced by the confederate and experienced anger over the nature of the questionnaire.

The interpretation of these results is that the emotional experience requires both autonomic arousal and a relevant cognition about the environment. Extending this notion further, it was proposed that the subjects explained their autonomic arousal by attributing it to the anger about the questionnaire, as expressed by the confederate. Since the questionnaire alone was a rather mild stimulus, it could not provide a sufficient account for the autonomic arousal until the flame was fanned, so to speak, by the confederate.

The explanation outlined above would be very tenuous, were it not for the complementary results of additional experiments. One such experiment used exactly the same treatments (adrenaline or placebo) and the subjects were asked to fill out a long and tedious questionnaire. This time, the confederate rebelled against the tedium of the task and began a high spirited game of basketball, using the wastebasket and some extra copies of the questionnaire. The pattern of results was the same: There was no particular emotion attached to the questionnaire per se for either the placebo or the adrenaline groups. Likewise, those subjects who had received the placebo paid little attention to the confederate. However, those subjects whose sympathetic nervous systems had been aroused by the adrenaline were strongly influenced by the antics of the confederate as revealed by their post-test expressions of euphoria.

There is a great deal of power in these two experiments. A particular emotion cannot be ascribed to the effects of a drug, the adrenaline. Nor can an emotional experience be triggered by the mere presence of a mild environmental situation. But the combination of sympathetic arousal and an appropriate environmental situation can produce a full blown emotional reaction. In the words of Schachter and Singer, the subjects who have been injected with adrenaline have a state of arousal that is in search of an appropriate cognition. These results have been extended in a number of novel designs, including one in which prior adrenaline injections increased the number of belly laughs during a slapstick comedy film. The framework of this theory has even included a naturalistic setting in which male subjects who had just walked across a high suspension bridge (presumably providing their own adrenaline) rated a female confederate significantly more attractive than males who had not crossed the bridge.

A final experimental manipulation provided the capstone for this notion of emotional experience. If the subjects were informed that the drug that they had received was adrenaline and told that it would produce an increase in heart rate, some flushing of the skin, and a general feeling of arousal, the emotional experience was forestalled: The symptoms were attributed to the drug action rather than to the antics of the confederate or the humor of the comedy.

The results of these experiments add a new dimension to the effects of various drugs, especially those that are designed to stabilize emotions or reduce anxiety. It is clear that the effects of these drugs could be either on the central interpretation of the environment (i.e., the cognition) or on the peripheral arousal aspects. It is very likely that the autonomic stabilizing effects play an important role in changing an individual's interpretation of the environment. Just as the subjects in Schachter's experiment say, in essence, that they must be experiencing an emotion because that is the only explanation they have for their state of arousal, so is it possible that an individual whose autonomic nervous system has been stabilized by an antianxiety agent may conclude that the situation must not be anxiety provoking because there is no autonomic arousal. Figure 4-16 shows a summary of some of these effects.

F. SUMMARY

Principles

1. A variety of different experimental procedures have been developed as models of situations that produce fear or anxiety in the natural environment. These procedures systematically vary the exposure to aversive events, the stimuli that signal these events, and the role of behavior in changing these relationships.

2. Pavlovian conditioning procedures show that fear can be evoked by previously neutral stimuli that have been paired with aversive events.

3. Instrumental conditioning involves two factors: Pavlovian conditioning of fear responses and learning of behaviors that are instrumental in changing these relationships.

4. The major response to short term stressors is the so-called flight or fight response of the sympathetic nervous system.

5. Longer exposures to stressors can result in the progressively more severe stages of the General Adaptation Syndrome.

6. Acute trauma such as surgery can lead to the shock syndrome, a diffuse outpouring of the entire autonomic nervous system.

7. The lack of a coping response for acute, profound stressors can lead to sudden death through overreaction of the parasympathetic nervous system.

8. The response to stress can be systematically changed by behavioral and pharmacological interventions.

9. The major forces that lead to ulcers are the inability to predict or control aversive events, and the presence of conflicting consequences (sometimes rewarded; sometimes punished) of behavior.

10. The stress response is more closely related to the interpretation of the environment than to the physical intensity of the aversive stimuli.

11. The search for better stabilizers of the autonomic nervous system led to the discovery of chlorpromazine and related phenothiazines known, collectively, as tranquilizers or antipsychotic drugs.

12. The benzodiazepines rather specifically reduce the effects of punishment, and are widely prescribed (e.g., Librium and Valium) as antianxiety drugs.

13. The phenothiazines have a high affinity for dopamine receptors.

14. The benzodiazepines have a high affinity for specific receptors that have not been linked to the GABA receptor complex. The presence of these receptors has suggested the possibility of an endogenous antianxiety compound in the brain.

15. Anticholinergic drugs appear to have excellent anti-punishment properties in animal experiments, but because of the peripheral side effects, they have little clinical value in the treatment of day to day anxieties.

16. The quaternary forms of atropine and scopolamine have been useful experimentally because they block cholinergic synapses in the periphery, but do not cross the blood brain barrier.

17. Cimetidine (Tagamet) is a very specific blocker of the H2 histamine receptor, and is widely prescribed to reduce the gastric acid secretion that can lead to ulcers.

18. Feedback from the autonomic nervous system plays an important role in determining whether or not an emotion will be experienced; environmental cues interact with this feedback to determine the nature of the emotional response.

Terms

Alarm reaction

Anticholinergics

Antihistamines

Atropine

Avoidance learning

Benzodiazepines

Blood brain barrier

Catecholamines

CER

Chlorpromazine

Cimetidine

Clinical potency

Conflict

CR

CS

Curare

Delay conditioning

Dopamine

Epinephrine

Escape learning

GABA

General Adaptation Syndrome

Generalized fears

H1 receptors

H2 receptors

Haloperidol

Histamine

James-Lange Theory

Long delay conditioning

Methyl atropine

Methyl scopolamine

Nicotine

Norepinephrine

Peptides

Phenothiazines

Placebo

Punishment

Receptor binding

Scopolamine

Stage of exhaustion

Stage of resistance

Sudden death

Surgical shock

Trace conditioning

Triad design

Two way avoidance

Two factor theory

Ulcers

UR

US

Yoked control