About the AuthorDavid H. Barlow received his PhD from the University of Vermont in 1969 and has pub-lished over 400 articles and chapters and over 20 books, mostly in the area of anxietydisorders, sexual problems, and clinical research methodology. Dr. Barlow was formerly Professor of Psychiatry at the University of Mississippi Medi-cal Center and Professor of Psychiatry and Psychology at Brown University, and foundedclinical psychology internships in both settings. He was also Distinguished Professor in theDepartment of Psychology at the University at Albany, State University of New York atAlbany. Currently, he is Professor of Psychology, Research Professor of Psychiatry, Direc-tor of Clinical Training Programs, and Director of the Center for Anxiety and Related Dis-orders at Boston University. Dr. Barlow is the recipient of the 2000 American Psychological Association (APA)Distinguished Scientific Award for the Applications of Psychology. He is also the recipientof the First Annual Science Dissemination Award from the Society for a Science of ClinicalPsychology of the APA, and recipient of the 2000 Distinguished Scientific ContributionAward from the Society of Clinical Psychology of the APA. He received an award in ap-preciation of outstanding achievements from the General Hospital of the Chinese People’sLiberation Army, Beijing, China, with an appointment as Honorary Visiting Professor ofClinical Psychology. Other awards include Career Contribution Awards from the Massa-chusetts and California Psychological Associations and a MERIT award from the NationalInstitute of Mental Health for long-term contributions to the clinical research effort. He isPast-President of the Society for Clinical Psychology and the Association for Advancementof Behavior Therapy, and is currently Editor of the Journal Clinical Psychology: Scienceand Practice. He was also a member of the DSM-IV Task Force of the American Psychiat-ric Association and was Co-Chair of the Work Group for revising the anxiety disordercategories. He is also a Diplomate in Clinical Psychology of the American Board of Profes-sional Psychology and maintains a private practice. vii
Contributing AuthorsMartin M. Antony, PhD, Anxiety Treatment and Research Centre, St. Joseph’s Hospital,Hamilton, Ontario, Canada; Department of Psychiatry and Behavioural Neurosciences,McMaster University, Hamilton, Ontario, CanadaTimothy A. Brown, PsyD, Center for Anxiety and Related Disorders, Department of Psy-chology, Boston University, Boston, MATerence M. Keane, PhD, National Center for PTSD in Boston; Boston Veterans Adminis-tration Healthcare System; Departments of Psychiatry and Psychology, Boston University,Boston, MAStefan G. Hofmann, PhD, Assistant Professor of Psychology, Center for Anxiety andRelated Disorders, Department of Psychology, Boston University, Boston, MASusan M. Orsillo, PhD, Boston Veterans Administration Healthcare System; Center forAnxiety and Related Disorders, Department of Psychology, Boston University, Boston, MALizabeth Roemer, PhD, University of Massachusetts at Boston; Center for Anxiety andRelated Disorders, Department of Psychology, Boston University, Boston, MAGail Steketee, PhD, School of Social Work, Boston University, Boston, MAKamila S. White, PhD, Center for Anxiety and Related Disorders, Department of Psychol-ogy, Boston University, Boston, MA ix
PrefaceTo provide some background on the genesis of this book, first published in 1988, it seemsworthwhile to revisit a few paragraphs from the preface to the first edition. At that time Iobserved: “This tale grew in the telling,” as J.R.R. Tolkein put it. After years of research and practice, it seemed a straightforward enough task to write a book on the nature and treatment of anxiety disorders. The surge of interest in these disorders during the past 5 years sparked rapid-fire advances in both psychological and pharmacological treatments, and I felt that such a book would be timely. After reviewing these advances, I planned to describe newly developed treat- ments for the anxiety disorders at our Center for Stress and Anxiety Disorders in Albany that take into account innovations from around the world. These treatments would be described in the framework of anxiety disorders as defined and described in the revisions to DSM-III. I told my publisher that I would have it for him in 12 to 18 months. During this time, our own treatments changed dramatically as exciting new discoveries emerged from our research center and elsewhere on the nature and treatment of panic. These developments seemed to require, at the very least, a full explication of the nature of anxiety and panic based on these new developments. But these developments could not be described adequately without putting them in the framework of current theoretical conceptions of anxi- ety in general. I found that repeating tried and true theories of anxiety proved totally inade- quate, since new conceptual approaches, ranging from neurobiological to social-constructivist, have appeared only in the past 2 years. These models are beginning to affect our view of anxi- ety. And then there is the mystery of panic! This confusing combination of new and old perspectives on anxiety and panic led inexora- bly to a consideration of the nature of emotion. The book stood still for a half a year while I absorbed once again the views of emotion theorists and attempted to integrate our rapidly emerging clinical knowledge of anxiety and panic with the old and distinguished tradition of emotion theory dating back to Darwin over 100 years ago. For anxiety disorders, in the last analysis, are emotional disorders. It became increasingly clear that it was impossible to say something fresh about emotional disorders without considering the traditions of emotion theory. At this point, I started working backward, rewriting the book from the point of view of the accumulated wisdom of emotion theory. After the integration of relevant recent developments in cognitive science and neuroscience, what emerged was a new model of panic and anxiety with implications for treatment. These theories and concepts were then integrated with newly developed treatment protocols for the various anxiety disorders. Only then was I able to write the book I intended. In writing this second edition I had the good fortune to be subjected to some of myown illusions of control. First, with the prospect of revising the book rather than writing xi
xii Prefaceit anew, and with the privilege of spending the 1997–1998 year at the Center for AdvancedStudies in Behavioral Sciences in Palo Alto, California, I thought I could finish the book ina year and a half and have it to my publisher in the late fall of 1998. As with the firstedition, I was off by 2 years! The explosion of knowledge in epidemiology and phenom-enology, as well as in cognitive, behavioral, and neurological science, made it clear that Iwould be lucky to finish even the first half of the book in 3 years’ time. With my closest colleagues at Boston University and the Center for Anxiety and Re-lated Disorders (CARD), where I relocated in the fall of 1996, we devised a plan that inthe second edition we would collaborate on the chapter on classification of anxiety disor-ders, as well as the last six chapters devoted to the nature, assessment, and treatment ofeach individual anxiety disorder. Those chapters apply the theory explicated in the firstpart of the book. Since all of the collaborating individuals are either current colleagues atBoston University and CARD, or in the case of Martin M. Antony, a former (stellar) stu-dent who is now an international authority in his own right, a full integration and conti-nuity of the ideas throughout the book was ensured. As I sit here putting the finishing toucheson this massive effort, it seems very possible to me that I may owe my own sanity to mygood friends and close collaborators, each of whom has achieved a more important mas-tery of his or her individual topic than I would have ever hoped to do alone. To TimothyA. Brown, Kamila S. White, Martin M. Antony, Terence M. Keane, Stefan G. Hofmann,Lizabeth Roemer, Susan M. Orsillo, and Gail Steketee, my deepest and most sincere grati-tude. The book is better for their efforts. There is one less chapter in this edition (15) than in the first edition, but there havebeen conceptual and structural changes in three chapters. Given enormous advances in ourknowledge of the process and nature of anxiety, this topic is now accorded its own chap-ter (Chapter 3). This chapter follows Chapter 1, on the experience of anxiety, which isconsiderably rewritten and updated, and Chapter 2, on fear, anxiety, and theories of emo-tion, which summarizes the burgeoning literature on emotion theory as it may apply toemotional disorders. Chapters 4, 5, and 6, on the phenomenon of panic, provoking panicin the laboratory, and neurobiological aspects of anxiety and panic, respectively, remainunchanged in title but enormously changed in content. Chapters 7 and 8 integrate the pre-ceding information and update my theoretical approaches to the origins of fear and panicand the origins of anxiety, respectively. To accommodate all of this additional information, the first edition’s Chapter 8, onthe process of fear and anxiety reduction, had to be dropped, although many of its ideason applying the principles of emotion theory to our new psychological and pharmacologi-cal interventions can be found in the later clinical chapters of this new edition. After Chap-ter 9, which covers classification, the last six chapters are devoted to the individual clini-cal anxiety disorders, including and combining two chapters from the first edition, on panicdisorder, into one. Much of what was hypothesized in the first edition concerning the nature and treat-ment of anxiety has been borne out in the ensuing 14 years, while other suggestions haveproven considerably off-base. During the 1990s, the centrality of emotion theory as a le-gitimate basis for the investigation of emotional disorders was recognized, and the studyof emotion from a variety of perspectives has been reinvigorated. The uniqueness of thephenomenon of panic, first recognized by Donald Klein, has been reaffirmed vis-à-vis therelated but distinct emotion of anxiety, and the functional relationship of these two emo-tions, first suggested in the first edition of this book, have become the basis for some of thecriteria for anxiety disorders in DSM-IV. The provocation of panic in the laboratory, asubject of considerable controversy at the end of the 1980s, has been confirmed by leading
Preface xiiineurobiological investigators to be mediated largely by psychological factors in some re-cent seminal studies; this also confirms a suggestion made in the first edition. And the senseof a lack of control, rooted deep in early experience, as the core of anxiety has receivedincreasing support in both animal and human models. On the other hand, conceptions of the process of worry postulated in the first editionhave changed considerably, based on the pioneering work of Thomas Borkovec, RudolphHoehn-Saric, and their colleagues as detailed in Chapter 3. Advances in the last decadehave also altered my opinions on the classifications of anxiety and related disorders, as Ihave recognized the overarching importance of broad dimensions of negative affect andneuroticism. From the point of view of treatment, the role of interoceptive exposure, ini-tially described in the first edition of this book, has broadened and deepened considerably.We also approach situational exposure in some important new ways. Many readers were intrigued with my dedication in the first edition to my children,then teenagers, a dedication I have altered only slightly in this edition to happily accom-modate our son-in-law and daughter-in-law. The “illusion of control” mentioned thererefers, of course, to a fundamental psychological characteristic that contributes substan-tially to physical and mental health, an idea that has received increasing support and isdescribed in some detail in Chapter 8. I am deeply appreciative of comments I have re-ceived from readers, not only on the dedication but all aspects of the book. Many of theseconcepts have assisted me greatly in advancing my own thinking. Since my own style of research has always been collaborative, this book owes muchto all of my colleagues and students over the years who have worked with me to developthese ideas. I take full responsibility for the fact that the ideas and thoughts of those indi-viduals may not have been represented exactly as they intended. Nevertheless, in most casestheir contributions will be recognizable. A special thanks to Michelle Craske, my colleague for over 15 years, who continuesto amaze me with her capacity for work and creativity. Thanks also to Sue Mineka andMark Bouton, with whom I spent a very enjoyable year at the Center for Advanced Stud-ies. I can only say that the Center was as heavenly as we all thought it would be. My thanksagain to Janet Klosko and Patty Coakley, who made important contributions to the moreliterary initiatives attempted in Chapter 1. Beyond these individuals, other colleagues onthe faculty here at Boston University have provided me with the benefit of their thinking,including David Spiegel, Henry Marcucella, Fabio Idrobo, and Kim Saudino. To my ad-ministrative assistants Bette Selwyn, who began the book, and Rebekah Morris, who dem-onstrated her uncanny ability to find missing references while finishing the book, I expressmy admiration and deepest appreciation. To my wife, Beverly, enduring the nights, week-ends, and “vacations” I spent working on the book, my gratitude and love for her patience.And to my copyeditor, Marie Sprayberry, who saw and considered carefully every wordof the first edition, and now has accomplished the same task with the second edition, myappreciation for her remarkable talent and attention to detail. And finally, as I bring this edition to a close with no expectations whatsoever ofattempting this Herculean task again, I look with enthusiasm to the next generation of in-vestigators who will share my excitement in continuing to unravel the mysteries of anxiety. David H. Barlow
ContentsCHAPTER 1 The Experience of Anxiety: Shadow of Intelligence 1 or Specter of Death?CHAPTER 2 Fear, Anxiety, and Theories of Emotion 37CHAPTER 3 The Nature of Anxious Apprehension 64CHAPTER 4 The Phenomenon of Panic 105CHAPTER 5 Provoking Panic in the Laboratory 139CHAPTER 6 Biological Aspects of Anxiety and Panic 180CHAPTER 7 True Alarms, False Alarms, and 219 Learned (Conditioned) Anxiety: The Origins of Panic and PhobiaCHAPTER 8 The Origins of Anxious Apprehension, 251 Anxiety Disorders, and Related Emotional Disorders: Triple VulnerabilitiesCHAPTER 9 Classification of Anxiety and Mood Disorders 291 Timothy A. Brown and David H. BarlowCHAPTER 10 Panic Disorder 328 Kamila S. White and David H. BarlowCHAPTER 11 Specific Phobias 380 Martin M. Antony and David H. BarlowCHAPTER 12 Posttraumatic Stress Disorder 418 Terence M. Keane and David H. Barlow xv
xvi ContentsCHAPTER 13 Social Phobia (Social Anxiety Disorder) 454 Stefan G. Hofmann and David H. BarlowCHAPTER 14 Generalized Anxiety Disorder 477 Lizabeth Roemer, Susan M. Orsillo, and David H. BarlowCHAPTER 15 Obsessive–Compulsive Disorder 516 Gail Steketee and David H. BarlowReferences 551Author Index 677Subject Index 694
C H A P T E R 1The Experienceof AnxietyShadow of Intelligence or Specter of Death?In our society, individuals spend billions of dollars yearly to rid themselves of anxiety. Thecosts of visits to primary care physicians, and the utilization of health care services in generalby individuals with anxiety disorders, are double what they are for those without anxietydisorders, even if the latter are physically ill (Simon, Ormel, Von Korff, & Barlow, 1995).Many take away prescriptions from their physicians for various drugs to treat anxiety, mak-ing these drugs among the most widely used in the world. For these individuals, anxiety is acurse—something they could live without. But could we all live without anxiety? Many ofour most prominent philosophers, psychologists, and psychiatrists think not. Some think thatit serves a protective function. For others, it is at the very root of what it means to be human.Still others believe that our very ability to adapt and plan for the future depends on anxiety.Consider the thoughts of the well-known early psychologist Howard Liddell: The planning function of the nervous system, in the course of evolution, has culminated in the appearance of ideas, values, and pleasures—the unique manifestations of man’s social living. Man, alone, can plan for the distant future and can experience the retrospective pleasures of achievement. Man, alone, can be happy. But man, alone, can be worried and anxious. Sherrington once said that posture accompanies movement as a shadow. I have come to be- lieve that anxiety accompanies intellectual activity as its shadow and that the more we know of the nature of anxiety, the more we will know of intellect. (Liddell, 1949, p. 185) Is anxiety the shadow of intelligence? Or is anxiety the overwhelming specter of deathand nothingness? The hypothetical advantages of being anxious are considered first.THE SHADOW OF INTELLIGENCESurvivalSince women reportedly suffer disproportionately from the pathological effects of anxi-ety, it is fitting that one of the most eloquent and lucid portrayals of the experience ofanxiety comes from a woman. In a short story by M. F. K. Fisher entitled “The Wind Chill 1
2 ANXIETY AND ITS DISORDERSFactor,”1 the character Mrs. Thayer is staying alone in a friend’s cottage on the ocean duringa brutal winter blizzard. The blizzard has been raging for days; nevertheless, Mrs. Thayeris warm, comfortable, and unconcerned with possible consequences of the storm as shedrifts off to sleep. A little after four, an extraordinary thing happened to her. From deep and comfortable dreamings she was wrenched into the conscious world, as cruelly as if she had been grabbed by the long hairs of her head. Her heart had changed its slow, quiet beat and bumped in her rib cage like a rabbit’s. Her breath was caught in a kind of net in her throat, not going in and down fast enough. She touched her body and it was hot, but her palms felt clammy and stuck to her. Within a few seconds she knew that she was in a state—perhaps dangerous—of pure panic. It had nothing to do with physical fear, as far as she could tell. She was not afraid of being alone, or of being on the dunes in a storm. She was not afraid of bodily attack, rape, all that. She was simply in panic, or what Frenchmen home from the Sahara used to call le cafard affolé. (p. 162)At this point, Mrs. Thayer experiences an urge to flee that is among the most fundamentaland ancient in the behavioral repertoire of living organisms. This is amazing, she said. This is indescribable. It is here. I shall survive it or else run out howling across the dunes and die soon in the waves and the wind. Such a choice seemed very close and sweet, for her feeling was almost intolerably wishful of escape from the noise. It was above and against and around her, and she felt that it was invading her spirit. This is dangerous indeed, she said, and I must try not to run outside. That is a suicide wish, and weak. I must try to breathe more slowly, and perhaps swallow something to get back my more familiar rhythms. She was speaking slowly to her whole self, with silent but precise enunciation. (p. 162) Is this experience Liddell’s shadow of intelligence? Is there anything intelligent or evenuseful in an unbearably strong urge to flee the warmth and security of a safe shelter to runheadlong into a raging sea? Paradoxically, there probably is, and this tendency may wellhave been responsible for the survival of the species. But is it only humans who can expe-rience “anxiety” in this way? According to the naturalist Charles Darwin, certainly not. With all or almost all animals, even with birds, Terror causes the body to tremble. The skin becomes pale, sweat breaks out, and hair bristles. The secretions of the alimentary canal and of the kidneys are increased, and they are involuntarily voided, owing to the relaxation of the sphincter muscles, as is known to be the case with man, and as I have seen with cattle, dogs, cats, and monkeys. The breathing is hurried. The heart beats quickly, wildly, and violently; but whether it pumps the blood more efficiently through the body may be doubted, for the surface seems bloodless and the strength of the muscles soon fails. . . . The mental faculties are much disturbed. Utter prostration soon follows, and even fainting. A terrified canary-bird has been seen not only to tremble and to turn white about the base of the bill, but to faint; and I once caught a robin in a room, which fainted so completely, that for a time I thought it was dead. (Darwin, 1872, p. 77)1Fisher, M. F. K. The wind chill factor or, a problem of mind and matter. In As They Were by M. F. K. Fisher.Copyright 1982 by M. F. K. Fisher. Reprinted by permission of Alfred A. Knopf, Inc. Originally appeared inThe New Yorker. Page numbers of excerpted passages refer to those in S. Cahill (Ed.), Women and Fiction 2:Short Stories by and about Women (pp. 160–166). New York: New American Library, 1978.
The Experience of Anxiety 3 Darwin may have developed a keen interest in these responses for a very good reason.It now seems clear that Darwin himself suffered the ravages of terror in the form of panicattacks, and that these attacks and the resulting agoraphobia kept him close to home,working with an all-consuming passion on his most famous works On the Origin of Spe-cies by Means of Natural Selection and The Expression of the Emotions in Man and Ani-mals (Barloon & Noyes, 1997). How can this elemental response be useful? As every student in introductory psychol-ogy knows, this behavior and the overwhelming emotion associated with it represent theorganism’s alarm reaction to potentially life-threatening emergencies. The almost reflex-ive urge to escape or, alternatively, to stand and engage the threat (“fight or flight”) seemsclearly a behavioral tendency that has been selectively favored in an evolutionary sense.Organisms without this capacity undoubtedly were overwhelmed by the welter of emer-gencies when the species was young. Organisms able to respond quickly and efficiently tolife-threatening situations survived and won the day. On close inspection, the specific functional contributions of various components ofthis biological survival mechanism seem clear (Cannon, 1927). Activation of the cardio-vascular system is one of the major components. Typically, peripheral blood vessels con-strict, thereby raising arterial pressure and decreasing blood flow to the extremities. Ex-cess blood is redirected to the skeletal muscles that can be used to defend oneself in a struggle.Blood pooled in the torso is more available to vital organs that may be needed in an emer-gency. Often people seem “white with fear”; that is, they blanch with fear as a result ofdecreased blood flow to the skin. Trembling with fear may be the result of shivering andperhaps piloerection, in which body hairs stand erect to conserve heat during periods ofvasoconstriction. These defensive adjustments can also produce the commonly observed“hot” and “cold” spells. Breathing becomes more rapid and usually deeper to providenecessary oxygen to rapidly circulating blood. This increased blood circulation carriesoxygen to the brain, where cognitive processes and sensory functions are stimulated. Anincreased amount of glucose (sugar) is also released from the liver into the bloodstream tofurther energize various crucial muscles and organs, including the brain. Pupils dilate, pre-sumably to allow a better view of the situation. Hearing becomes more acute, and diges-tive activity is suspended, resulting in a reduced flow of saliva (the “dry mouth” of fear).There is often pressure to urinate, to defecate, and occasionally to vomit. In the short term,voiding will further prepare the organism for concentrated action and activity; in the longerterm, vomiting and diarrhea may be reflexive reactions protecting against the danger ofabsorption of noxious substances (Beck & Emery with Greenberg, 1985; Nesse, 1987; Stein& Bouwer, 1997). Is this mobilization of the organism for fighting or escaping the only behavior associ-ated with our ancient alarm system? It seems not. Consider the following case from ourfiles: A young physician in radiology was attending an Ingemar Bergman movie one evening.The camera focused suddenly on massive bleeding from one character in an unexpectedcontext. The physician slumped over. His companion thought he might be napping, al-though he had seemed quite interested in the movie to that point. When the physician cameto, he was very shaken, realizing that he had fainted unexpectedly. Even more disconcert-ing was the fact that his faint was due to the sight of blood. In fact, throughout medicalschool, he had often noticed feeling uneasy in the presence of blood, but had “steeled”himself to the occasion. In all likelihood, he had employed various muscle-tensing strate-gies that maintain or increase blood pressure (Öst & Sterner, 1987; Antony, Craske, &Barlow, 1995b; Craske, Antony, & Barlow, 1997) (see Chapter 11). In any case, he re-mained sufficiently uncomfortable to choose a specialty where contact with blood was
4 ANXIETY AND ITS DISORDERSminimized. After the Bergman film, his discomfort escalated to a full phobic reaction, caus-ing the avoidance of any situation where he might encounter blood. Fainting, of course, requires a very different physiological response from that requiredby fighting or fleeing. Instead of the sustained sympathetically innervated surges in car-diovascular function that are associated with the usual alarm reaction, marked decreasesin heart rate and blood pressure precede fainting (Page, 1994). Could this also be an adap-tive response? Very clearly it could, and it probably was such a response under conditionsthat existed in millenia past. When under attack, those who responded to injury and bleedingwith a dramatic drop in blood pressure—thereby minimizing blood loss and the danger ofshock—were far more likely to survive the attack than those who did not. Today this be-havioral action tendency is maladaptively present in the large numbers of individuals withextreme fears of blood, injury, and injection (see Chapter 11). But what of Darwin’s robin, which presumably did not see blood, but neverthelessfainted so completely when caught that Darwin thought it was dead? Is this the same re-action as that described above? Probably not. In recent years, investigators have rediscov-ered an archaic response that seems to be yet another action tendency associated with alarmreactions in specific situations. When in the presence of an approaching predator, andparticularly when in direct contact with a predator, most species will initially evidence theagitation associated with fight or flight. This reaction may be followed immediately by avery different response similar to paralysis, but characterized by waxy flexibility. Theseanimals look as if they were dead, as did Darwin’s robin. In fact, this is the “playing dead”or “freezing” response so often seen in animals in the wild under attack. Investigators inthis area refer to the response as “tonic immobility” (Gallup, 1974) and differentiate itfrom the temporary motionless response exhibited by many animals preceding their at-tempts to flee from a predator seen at a distance (Woodruff & Lippincott, 1976). But in-vestigators have determined that tonic immobility is not a volitional or strategic strategyon the part of the animal. Rather, this response represents another ancient behavioral re-action with obvious survival value. For the large number of predators for which attack istriggered and maintained by movement, freezing is an effective antidote that prevents fur-ther attack and increases the victim’s chances of survival. This response may have important biological implications for human anxiety, as out-lined in Chapter 6. But investigators concerned with human anxiety have showed littleinterest in tonic immobility, since it has not been thought to occur in humans. Now itseems we may have overlooked a tragic but obvious example of this reaction in humans.During brutal rapes, many women report feeling paralyzed. Comments such as “I felttrembling and cold—I went limp,” or “My body felt paralyzed,” or “My body wentabsolutely stiff,” reflect this paralytic state. This may be analogous to the muscular ri-gidity and motor inhibition evidenced by animals during tonic immobility followingmanual restraint. Loss of consciousness does not occur, since the victim can later relateevents that occurred during the attack. Some victims have reported feeling “freezing cold,”which may reflect the characteristic decrease in body temperature of immobile animals.Many survivors of rape also report feeling completely numb or insensitive to pain dur-ing the ordeal; this may be similar to the analgesic effects of immobility observed inanimals. Burgess and Holmstrom (1976) reported that 22 out of 34 rape survivors dem-onstrated physical paralysis at some point during the encounter. The similarities betweentonic immobility and rape-induced paralysis have been outlined by Suarez and Gallup(1979) and are found in Table 1.1. It is not clear that the survival response of freezing obviates some of the dangers ofrape. One could speculate that freezing decreases the risk of injury from physical aggres-
The Experience of Anxiety 5 TABLE 1.1. Similarities between Tonic Immobility and Rape-Induced Paralysis Tonic immobility Rape-induced paralysis Profound motor inhibition Inability to move Parkinsonian-like tremors Body shaking Suppressed vocal behavior Inability to call out or scream No loss of consciousness Recall for details of the attack Apparent analgesia Numbness and insensitivity to pain Reduced core temperature Sensation of feeling cold Abrupt onset and termination Sudden onset and remission of paralysis Aggressive reactions at termination Attempts to attack the rapist following recovery Note. From Suarez and Gallup (1979). Copyright 1979 by The Psychological Record. Reprinted by permission.sion. Furthermore, restricted movement cues may reduce sexual arousal in the rapist (Suarez& Gallup, 1979). But recognizing that rape survivors are paralyzed or tonically immobilecould prevent a tragic interpretation by many authorities, who have assumed in the pastthat a survivor is somehow acquiescing to the rape.The Sound of the WindIf these responses are termed “anxiety,” and one considers it “intelligent” to avoid threatssuccessfully, then anxiety may be the shadow of intelligence. But most would not considerthis behavior to be “intelligent” in the sense of a complex rational response. Nevertheless,it shares something of Howard Liddell’s conception, in that it is adaptive, useful, and in-dispensable. But there is something very mysterious about the fictional Mrs. Thayer’s experiencethat does not fit the Darwinian scenario. “It had nothing to do with physical fear, as far asshe could tell. She was not afraid of being alone, or of being on the dunes in a storm. Shewas not afraid of bodily attack, rape, all that” (Fisher, 1978, p. 162). The threat in herown mind is not the storm or being alone, but the urge to flee itself: “This is dangerousindeed, she said, and I must try not to run outside. That is a suicide wish, and weak”(p. 162). But the “suicide wish” is more than a flirtation with death running through herhead in the form of frightening thoughts. There are other components to this threat ema-nating from within her own body: The sound of the wind, for her, had been going sideways exactly on a line with the far horizon of the Atlantic for days, nights—too long. It was in her bowels and suddenly they were loos- ened, and later, again to her surprise, she threw up. She told herself dizzily that the rhythm of the wind had bound her around, and that now she was defying it, but it kept on howling. (p. 165) How does one cope with this human experience where the threat is internal and theconsequence is death? With great difficulty, Mrs. Thayer struggles to walk to the kitchen,where she takes two aspirins and a mug of warm milk. Ritualistically, she remembers ad-vice that during periods of deep stress one should drink the liquid in three slow sips, wait5 minutes before taking three more sips, and so on. After the medicine and the rituals, sheturns to other, more psychological methods of coping:
6 ANXIETY AND ITS DISORDERS She pulled every trick out of the bagful she had collected during her long life with neurotics. She brushed her hair firmly, and all the while her heart kept ticking against her ribs and she felt so sick that she could scarcely lift her arm. She tried to say some nursery rhymes and the Twenty-third Psalm, but with no other result than an impatient titter. She sipped the dreadful sweet milk. She prayed to those two pills she had swallowed. (p. 163)But at the last she turns to the great arbiter of all human difficulties, one’s own reason. For a time, as the aspirin and the warm milk seemed to slow down her limitless dread (Dread of what? Not that the roof would fly off, that she was alone, that she might die . . . ), she made herself talk reasonably to what was pulling and trembling and flickering in her spirit. She was a doctor—or, rather, an unwitting bystander caught in some kind of disaster, forced to be cool and wise with one of the victims, perhaps a child bleeding toward death or an old man pinned under a truck wheel. She talked quietly to this helpless, shocked soul fluttering in its poor body. She was strong and calm. All the while, she knew cynically that she was nonexist- ent except in the need thrust upon her, and that the victim would either die or recover and forget her dramatic saintliness before the ambulance had come. “Listen to your breathing,” she said coolly. “You are not badly hurt. Soon you will feel all right. Sip this. It will make the pain go away. Lift your head now, and breathe slowly. You are not really in trouble.” And so on. Whenever the other part of Mrs. Thayer, the threatened part, let her mind slip back to the horror of an imminent breaking with all reason—and then, so then, out the door it would be final—the kindly stranger seemed to sense it in the eyeballs and the pulse as she bent over the body and spoke more firmly: “Now hold the cup. You can. I know you can. You will be all right.” (p. 164) Mrs. Thayer perseveres and later decides that the pills, helped by the warm drink, haveworked. She concludes that her mind has not failed her as she attempts to distract herselfwith the Twenty-third Psalm and convince herself that “she would never have run out likea beast, to die quickly on the dunes” (p. 165). Later she reflects on her experience. In another two hours everything was all right inwardly in her, except that she was lan- guid, as if she had lain two weeks in a fever. The panic that had seized her bones and spirit faded fast, once routed. She was left wan and bemused. Never had she been afraid—that is, of tangibles like cold and sand and wind. She was not afraid, as far as she knew, of dying either fast or slowly. It was, she decided precisely, a question of sound. If the storm had not lasted so long, with its noise so much into her, into her brain and muscles . . . If this had been a kind of mating, it was without joy. . . . And during the late afternoon while she dozed with a deep, soft detachment, the sound abated and then died, and she was lost in the sweet dream life of a delivered woman. (p. 166)Terms and MeaningsMrs. Thayer’s experience, and her methods of coping both during and after the experience,illustrate what is essential and mysterious about the human emotion we call anxiety.Is it really the wind that causes Mrs. Thayer’s panic? Or is the search for causes necessi-tated by some fundamental human quality that cannot let events such as this go un-explained? Is it better that she find a cause, even if incorrect, than that she not searchat all? In fact, speculation concerning cues has played a pivotal role in theorizing about anxietyover the years. Many early theorists, among them Kierkegaard and Freud, based defini-tions and distinctions of “fear” and “anxiety” on the presence or absence of cues. For
The Experience of Anxiety 7most, one important distinction became prominent: “Fear” was seen as a reaction to aspecific, observable danger, while “anxiety” was seen as a diffuse, objectless apprehen-sion. Theorizing about anxiety, then, involved a search for “hidden” cues that was notunlike Mrs. Thayer’s speculations on the sound of the wind. This distinction between fearand anxiety produced the rich theoretical framework that still underlies much of our think-ing on the development of psychopathology. Standard definitions of “fear” and “anxiety”in dictionaries and introductory psychology textbooks continue to refer to the presence orabsence of identifiable cues as the essential distinction. The ascendance of direct behav-ioral approaches to treating fears and phobias began to change that, however (Wolpe, 1958).Behavior therapy assumed that all anxiety has clear identifiable cues, although some cuesare more diffuse than others (e.g., patterns of light and dark).The distinction between “fear”and “anxiety” became blurred for many, and these terms have until recently been equatedin psychology and psychiatry. But terminology describing experiences such as Mrs. Thayer’s is more varied and con-fusing than the hypothetical distinction between fear and anxiety. Among many terms incommon use in the English language today are “anxiety, “fear,” “dread,” “phobia,”“fright,” “panic,” and “apprehensiveness.” Each of these terms is often qualified with suchwords as “acute,” “morbid,” “generalized,” “diffuse,” and so forth to provide differentshades of meaning. In addition, no student of anxiety reflecting on terminology can omitthe German word angst. Although difficult to translate, this word forms the basis for muchof our thinking about the role of anxiety in psychopathology, since it was the word usedby both Kierkegaard and Freud. For Kierkegaard, angst would be both “dread” and “anxi-ety”; however, sometimes one is used in English translations of Kierkegaard and some-times the other. For Freud, angst came to reflect the notion of anxiety without an identifi-able object. Rather, angst was a vague apprehension about the future (although thetheoretical significance focused on the present and the past). When anxiety had an object,Freud also preferred the word furcht (fear). Although “anxiety” is the word with whichwe are now familiar, Sir Aubrey Lewis (1980) suggests that a more precise translation ofangst would be “agony,” “dread,” “fright,” “terror,” “consternation,” “alarm,” or “ap-prehension.” Essentially, the word angst signifies a far more shattering emotion than theEnglish word “anxiety,” which is often used as synonymous with “concern.” As Lewis(1980) points out, the relevant root word passed down from Greek and Latin is angh, whichrefers literally in Latin to the concept of narrowness or constriction. Various derivatives ofthis root have evolved differently in different Western languages, as one can see by exam-ining the number of words in English with the angh root. Among these are “anxiety,”“anguish,” and “anger.” The profusion of meanings and flavors surrounding the key words angst, “anxiety,”and “dread”; the somewhat different usages in different languages; and the imprecisionresulting from translations of seminal works have all resulted in an understandable vague-ness surrounding the term “anxiety” in English. The short history of the usage of the termin psychopathology has produced even less precision. In recent years, “anxiety” has beenused to refer to emotional states such as doubt, boredom, mental conflict, disappointment,bashfulness, and feelings of unreality. Various cognitive deficits, such as lack of concen-tration, are also labeled “anxiety.” In addition, the term has been inextricably bound upwith the variety of terms describing depressive emotional states. The emergence of theo-retical and descriptive qualifiers (e.g., “unconscious,” “conscious,” “cognitive,” “somatic,”“free-floating,” “bound,” “signal”) produces further confusion. For this reason, the diffi-culty in settling on precise distinctions among the anxiety-related terms in English is notsurprising.
8 ANXIETY AND ITS DISORDERS This state of confusion has caused some to propose that we drop the word altogether,since it is so imprecise as to be unscientific (Sarbin, 1964). In fact, Sarbin and other socialconstructivists such as Hallam (1985) have proposed new ways of thinking about the con-cept. For example, Hallam has suggested that anxiety is essentially a lay construct thatcan refer to vastly different cognitive and somatic points of reference from person to per-son. For social constructivists, anxiety is best considered a metaphor. For the moment, Iforgo my own definitions of the crucial terms “anxiety,” “fear,” “panic,” “apprehension,”and so on; I elaborate on these terms at appropriate points in later chapters.Being and PreparingMrs. Thayer can find no reason for her panic until deciding, upon reflection, that the causeis the incessant sound of the storm. For this reason, many would say that her experience isone of anxiety rather than fear, although since the 1980s many would employ the term“panic” as she herself does. But what could be the purpose of this experience? If Darwin-ian fear and panic facilitate survival, what is the purpose of “anxiety” in the traditionalusage of the term, where there is nothing to fear? Philosophers preoccupied with a searchfor cues have often decided that there is something very valuable indeed in this experience,which may lead one, in an ironic way, to a greater sense of fulfillment and actualization. Kierkegaard (1844/1944) was one of the first to make this suggestion when he decidedthat the source of anxiety is deep within the individual. Anxiety, thought Kierkegaard, isrooted not just in a fear of death, but in a fear of nonexistence, nonbeing, or nothingness.Only through recognizing and confronting this fear of becoming nothing—only throughthe threat of dissolution of the self—can one truly discover the essence of being. Onlythrough this experience can one achieve a clear distinction of the self from other objects orfrom nonbeing. Other well-known theorists and clinicians have settled on a similar cause for diffuseand objectless anxiety. For example, Rollo May (1979) has proposed: [Anxiety is] the apprehension cued off by a threat to some value that the individual holds es- sential to his existence as a personality. The threat may be to physical life (a threat of death), or to psychological existence (the loss of freedom, meaninglessness). Or the threat may be to some other value which one identifies with one’s existence (patriotism, the love of another person, “success,” etc.). (p. 180)In the sense that this confrontation results in a higher level of existence and a greater ap-preciation of what it is to be alive, this may represent the shadow of intelligence, and thepurpose and meaning of anxiety. As Freud saw it, angst can be cued by activation of elemental threats to the child, whichare stored deep in memory and elicited in the adult by a variety of learned associations(Freud, 1926/1959; Michels, Frances, & Shear, 1985; Shear, Cooper, Klerman, & Busch,1993). In this sense, anxiety is related to the persistence of remembered danger situationsthat seemed real at an earlier stage of development. For example, the developmentallyimmature fears of castration or separation may be activated by the emergence of an asso-ciated wish or by the occurrence of a symbolically linked situation currently present in one’senvironment. Anxiety functions to warn of a potential danger situation and triggers therecruitment of internal psychological and/or external protective mechanisms. The institu-tion of effective psychological defense mechanisms serves the adaptive purposes of pro-tecting the integrity of the individual and allowing a higher and more mature level of func-tioning. Anxiety may also be adaptive in recruiting help from others when there is real
The Experience of Anxiety 9danger. Sometimes defensive reactions are inadequate and lead to symptom formation.These may include phobic or compulsive symptoms that are symbolically related to theunconscious wishes or fears that have generated the anxiety. Self-defeating aspects of anxietyare further elaborated below. Freud would view the raging storm and incessant noise of Mrs. Thayer’s blizzard asproviding sensory stimuli indicative of a real threat. Anxiety occurs in reaction to the pos-sibility of being overwhelmed by this threat and rendered helpless. The sense of ultimateseparation and isolation one can only experience alone in a blizzard may also elicit memo-ries of childhood fears of separation. For Mrs. Thayer, this dual challenge leads to theemergence of unmanageable levels of anxiety and a strong primitive urge to find somehuman contact. But even these popular and still current ideas probably do not capture Liddell’s (1949)meaning when he talked of anxiety as the shadow of intelligence. Although Liddell wastalking of human experience, his scientific explorations concerned the development ofpathological anxiety in animals. In his most famous experiments, he produced what cameto be called “experimental neurosis” (see Chapter 8). One consequence of experimentalneurosis is that animals become more vigilant concerning future threats. Liddell theorizedthat vigilance has positive consequences in addition to simply helping the animal to noticemore quickly the next threat to its well-being. He observed that vigilant animals seem tobe conditioned or to learn more easily. Vigilance, therefore, which Liddell supposed to bethe animal counterpart of anxiety, may produce more learning and therefore more intelli-gent animals. But it is the type of learning that is particularly important. The vigilant ani-mal, occupied as it is with future threat, is concerned with what is going to happen in theimmediate future. In a very elementary sense, the animal is planning for that future bytaking an orientation to the future best characterized by the question “What happens next?”.The planning function is apparent. In humans, this is extremely adaptive. Liddell suggestedthat effective planning for the future and the retrospective enjoyment of past achievementsare the means by which human beings construct culture. The capacity to experience anxi-ety and the capacity to plan are therefore two sides of the same coin. It is in this sense thatanxiety accompanies intellectual activity as its shadow. Anyone who has succeeded at any task, however small, has probably experienced someaspect of this fascinating and mysterious quality of anxiety. For we have known for al-most 100 years that our physical and intellectual performance is driven and enhanced bythe experience of anxiety, at least up to a point. In 1908, Yerkes and Dodson demonstratedthis in the laboratory by showing that the performance of animals on a simple task wasbetter if they were made “moderately anxious” than if they were experiencing no anxietyat all. Since that time, similar observations have been made concerning human performancein a wide variety of situations and contexts. Without anxiety, little would be accomplished.The performance of athletes, entertainers, executives, artisans, and students would suffer;creativity would diminish; crops might not be planted. And we would all achieve that idyl-lic state long sought after in our fast-paced society of whiling away our lives under a shadetree. This would be as deadly for the species as nuclear war. In summary, several centuries of thought from very diverse sources have emphasizedthe importance of anxiety to creativity, intelligence, and survival itself. But it is unlikely thatMrs. Thayer, trapped in her cottage on the ocean at the height of a storm, comes to considerit a growth experience. For her in her fictional setting, as well as for countless millions ofindividuals in the course of their everyday existence, it is a dramatic life-and-death strugglewith the ever-present possibility that death may win out. And there is evidence that deathdoes win out on occasion, as a result of the cumulative consequences of anxiety.
10 ANXIETY AND ITS DISORDERSTHE SPECTER OF DEATHThe Neurotic ParadoxIn 1950, O. Hobart Mowrer described a mystery: [It is] the absolutely central problem in neurosis and therapy. Most simply formulated, it is a paradox—the paradox of behavior which is at one and the same time self-perpetuating and self-defeating! . . . Common sense holds that a normal, sensible man, or even a beast to the limits of his intelligence, will weigh and balance the consequences of his acts: if the net effect is favorable, the action producing it will be perpetuated; and if the net effect is unfavorable, the action producing it will be inhibited, abandoned. In neurosis, however, one sees actions which have predominently unfavorable consequences; yet they persist over a period of months, years, or a lifetime. (p. 486) This paradox is seen daily in clinics all over the world. Consider the case of JohnMadden, the well-known American sports announcer and former professional footballcoach, who has written widely about his anxiety and uses it in a humorous mannerin several television commercials. Although Madden has overcome the stigma andembarrassment that would be keenly felt by any 64", 260-pound former football playerwhose business is to be tough and courageous, he has not overcome the anxiety itself.Rather than taking a few hours to fly from New York to San Francisco to announce thenext football game, he must spend the better part of his week on a train going acrossthe country. In fact, his fears are not limited to planes, but extend to all claustrophobicsituations. Although Madden was always tense in planes, he originally thought of his tension asa reaction to altitude, probably the symptoms of an inner ear infection. When he realizedthat his anxiety began before the plane took off, but after the stewardess closed the door,he questioned his previous diagnosis. One day while flying across the country, he experi-enced a particularly severe panic attack, left the plane at a stop halfway across the UnitedStates, and never flew again. John Madden and countless millions of other individuals suffering from anxiety-baseddisorders are well aware that there is little or nothing to fear in the situations they find sodifficult. Therefore, in Mowrer’s terms, Madden should have long since weighed the con-sequences of his acts and decided that, since flying is the safest way to travel, it would bein his best interest to fly in order to save himself time and help maintain his lucrative ca-reer. And yet he does not and cannot abandon his self-defeating behavior. One might saythat if Madden at least attempted to fly, he would learn something that does not seemamenable to the rational force of persuasion by either himself or others—namely, that fly-ing is safe. But we know from years of clinical and scientific experience that even forcedexposure to difficult situations does not always resolve the paradox. The self-defeating nature of anxiety and its consequences is dramatically elaboratedin psychoanalytic theory. Freud saw anxiety as the psychic reaction to danger. A situa-tion can be defined as dangerous if it threatens a person with helplessness in the face ofthreat. Dangers regarding the external world lead to realistic anxiety, and dangersto conscience result in moral anxiety; however, dangers surrounding the strength ofthe passions lead to neurotic anxiety. Neurotic anxiety originates from an inner instinc-tual wish that is associated with a reactivation of an infantile fear situation. The genera-tion of anxiety in any of these spheres leads to the institution of a defense mechanism.All forms of anxiety occur in normal individuals. In the process of development, indi-
The Experience of Anxiety 11viduals learn to modify and modulate the expression of anxiety from its most disruptiveintense form to an unnoticeable form called “signal anxiety.” Signal anxiety is imper-ceptible to the person experiencing it, and serves the sole purpose of rapidly and effi-ciently triggering a defensive reaction. Thus normal anxiety is limited in intensity andduration, and is associated with adaptive defenses. Anxiety is self-defeating or patho-logical when it is noticeable, intense, disruptive, and paralyzing, or when it triggers self-defeating defensive processes, also called “symptoms.” Phobic and obsessive symptomsare especially common in reaction to anxiety. These symptoms represent an insufficientattempt at warding off a danger situation, and typically incorporate elements of thedanger. For example, a dog phobia may develop in connection with the activation ofan infantile fear of castration. The fear is displaced, but the aggressive component isretained. In the world of Freud, we confront our infantile modes of psychological function-ing. Pathological anxiety emerges in connection with some of our deepest and darkestinstincts. Before Freud, the embodiment of good and evil and of urges and prohibitionswas conceived of as external and spiritual, usually in the guise of demons confrontingthe forces of good. Since Freud, we ourselves have become the battleground of these forces,and we are inexorably caught up in the battle, sometimes for better and sometimes forworse.MortalityIf anxiety, in the minds of some, is apprehension over confronting nothingness, considerthe consequences of severe anxiety, which result occasionally in physical destruction ordeath. Consider the man with fears of choking so severe that he consumes only strainedand blended food; the results may be malnourishment, loss of teeth, and eventually death.Or consider the woman with obsessive–compulsive disorder (OCD) who ritualisticallywashes and disinfects her arms and legs for the better part of her waking day; the resultsmay be massive abrasions, bleeding, and scabbing.Cardiovascular DiseaseWhereas self-defeating behavior associated with anxiety may occasionally produce death,the long-term consequences of anxiety itself may hasten the dreaded confrontation withnothingness. Long-term follow-up studies of both inpatients (Coryell, Noyes, & Clancy,1982) and outpatients (Coryell, Noyes, & House, 1986) found a greater-than-expectedmortality rate in patients with original diagnoses of anxiety disorders, particularly panicdisorder. This excess mortality rate was attributed primarily to cardiovascular disease andsuicide. Interestingly, excess mortality due to cardiovascular disease in these studies waslimited to males with panic disorder. Expected death rates for females with panic disorderfrom cardiovascular disease were within normal range. More recent prospective studies, with larger groups of males, confirm this frighteningfinding. In the first of these studies, Kawachi, Colditz, et al. (1994) examined over time alarge group of over 33,000 male health professionals between the ages of 42 and 77, toassess the relationship between anxiety and the risk of coronary heart disease. Data wereobtained from a large-scale study of normative aging. Men with the highest levels of pho-bic anxiety had a level of risk for fatal coronary heart disease three times higher than thatof men with lower levels of anxiety. More importantly, the relative risk was limited to menexperiencing sudden cardiac death as opposed to nonsudden coronary death. In fact, men
12 ANXIETY AND ITS DISORDERSwith the highest levels of phobic anxiety had a relative risk of sudden death six times thatof men with the lowest anxiety levels. In a second study, the relationship between symp-toms of anxiety and coronary heart disease in over 2,200 male community residents aged21 through 80 revealed an increased risk of fatal coronary heart disease that was almostfive times that of males with few or no anxiety symptoms. It is significant that this elevatedrisk ratio was present even after confounding variables (family history of heart disease,smoking, blood pressure, etc.) were controlled for. One possible mechanism by whichanxiety might influence cardiac function is decreased heart rate variability in anxious in-dividuals, which can result in ventricular arrhythmias (Kawachi, Sparrow, Vokonas, &Weiss, 1995; Zaubler & Katon, 1996). Findings suggesting the relationship between de-creased heart rate variability and anxiety are becoming more robust (Yeragani et al., 1995)and are discussed in more detail in Chapter 6.SuicideDeath by suicide is an event most often associated with depression. But Coryell et al.(1986) found that the frequency of suicide in patients with anxiety disorders was equalto the frequency in matched groups suffering from depression. Why would people whoare anxious kill themselves? Coryell et al. (1986) speculated that patients diagnosed withanxiety disorders may subsequently develop major depression or alcoholism as a com-plication. They thought it possible that earlier studies had overlooked the occurrence ofsuicide in anxious patients because they noticed only the subsequent complication ofalcoholism or depression. But if alcoholism or depression is a consequence of anxiety,than the long road to suicide may begin with anxiety. This intriguing early study wasnot widely noted. But a subsequent study, published by Myrna Weissman and her colleagues in the NewEngland Journal of Medicine, attracted international attention from public health com-munities, this study noted that, based on data from a community survey—the Epidemio-logic Catchment Area (ECA), study described in more detail below—some 20% of patientswith panic disorder had made a suicide attempt at some point during their lives (Weissman,Klerman, Markowitz, & Ouellette, 1989). This finding was alarming to clinicians whotreated panic disorder. Weissman et al. (1989) also agreed with Coryell et al. (1986) thatthe risks of suicide in patients with panic disorder were comparable to those associatedwith major depression. In a subsequent report (Johnson, Weissman, & Klerman, 1990),the investigators were very careful to analyze panic disorder either with or without comorbiddiagnoses such as depression or substance abuse, and found that the risks of suicide weresubstantially elevated even in patients without these cormorbid disorders. This finding then became very controversial, since clinicians had difficulty con-firming the presence of suicidal risk in patients with panic disorder under their care. Forexample, my colleagues and I (Friedman, Jones, Chernen, & Barlow, 1992) and Beck,Steer, Sanderson, and Skeie (1991) examined hundreds of patients with panic disorderin outpatient clinics, and did not find a suicidal risk that seemed to be associated exclu-sively with panic disorder. That is, any suicidal risks (e.g., suicidal attempts or ideation)seemed to be present only in those patients with additional disorders (such as comorbidborderline personality disorder, in which suicidal attempts are a prominent characteris-tic). For example, Beck et al. (1991) reported that in their last 900 outpatients with panicdisorder, all suicide attempts were attributed to a comorbid depressive disorder. Thiswas in direct contradiction to the large community study reported by Weissman et al.(1989).
The Experience of Anxiety 13 What accounts for these marked discrepancies? Of course, the method of assessmentand type of interview used in the studies differed considerably, as did the populations stud-ied. Individuals in the ECA study came from a random sample of the community. The clinicalsamples consisted of patients seeking treatment. Thus it is possible that the ECA study mayhave included individuals who not only were untreated for panic disorder, but also wereunaware of the possibility of treatment for this disorder. As a result, they might have be-come hopeless and begun to think about suicide. All patients at the clinics, on the otherhand, were properly diagnosed and about to receive treatment. On the other hand, Becket al. (1991) and our group (Friedman et al., 1992) also checked for lifetime suicide at-tempts in the clinical samples and did not find elevated risk at any time, which would in-clude periods prior to seeking treatment. Also, Hornig and McNally (1995) looked moreclosely at the original ECA data from Weissman and colleagues’ studies, and suggestedafter additional analyses that elevated risk of suicide was associated with comorbid diag-noses, if the data on comorbidity were analyzed in the aggregate rather than by individualcomorbid disorder. The discrepancy remains something of a mystery, but other studies have now appearedthat continue to alarm clinicians. For example, Noyes (1991) reviewed a number of earlystudies, many examining the relationships of suicide attempts to anxiety disorders in gen-eral, and concluded there was an elevated risk. In a large and important study, Allgulanderand Lavori (1991) examined over 3,000 inpatients in Sweden with “pure” anxiety neuro-sis (i.e., without comorbid depressive or other diagnoses). They concluded that the risk ofcompleted suicide in these patients was as high as that in persons with depression. In afollow-up study, Allgulander (1994) expanded his inquiry to almost 10,000 patients withanxiety disorders and concluded that the risk for completing suicide before the age of45 years among men and women with anxiety disorders without any other psychiatricdiagnoses was between 4.9 and 6.7 times that of the risk in the general population. Ex-amining the issue from another perspective, Norton, Rockman, Luy, and Marion (1993)noted that patients with substance use disorders and accompanying panic disorder wereat substantially more risk for attempting suicide than those individuals with substanceuse disorders without panic disorder—a finding also reported by Hornig and McNally(1995). Although many questions remain, and further analyses are required, evidence isconverging that the experience of anxiety and panic may increase the probability of end-ing one’s life.Anxiety and Substance Abuse/DependenceSuicide may be an extreme consequence of the experience of anxiety, but evidence indi-cates that another type of self-destructive behavior is not. A series of investigations sug-gests that the relationship of substance use disorders, particularly alcohol abuse and de-pendence, to anxiety disorders is startingly high. In an early study, Quitkin, Rifkin, Kaplan,and Klein (1972) reported in some detail on 10 patients with anxiety disorders who alsosuffered severe complications from drug and alcohol dependence. The important sugges-tion by Quitkin et al. (1972) was that patients presenting with substance dependence maywell be self-medicating an anxiety disorder. Thus any treatment program for individualswith alcohol or other substance use disorders must target this anxiety if the program is tobe successful, as well as if it is to prevent relapse. Although this has always been a common clinical observation, a study by Mullaneyand Trippett (1979) attracted new attention to this potentially severe complication ofanxiety. They discovered that 33% of 102 individuals with alcohol dependence also had
14 ANXIETY AND ITS DISORDERSsevere, disabling agoraphobia and/or social phobia. In addition, another 35% had mildversions of the same phobias. Thus over 60% of a large group of patients admitted to analcoholism treatment unit presented with identifiable anxiety disorders of varying severity(in this case, phobias). Smail, Stockwell, Canter, and Hodgson (1984) explored this samequestion in a particularly systematic and skillful way. They found similar although some-what less dramatic results. Specifically, 18% of a group of 60 patients presenting withalcohol dependence also presented with severe agoraphobia and/or social phobia, withanother 35% evidencing mild versions of the same phobias. Once again, over half (53%)of a group of individuals dependent on alcohol had identifiable phobic disorders. Since that time, studies using even more conservative criteria have also reported a ratherhigh range from 25% to 45% of patients with severe alcohol use problems presenting withone or more anxiety disorders (e.g., Chambless, Cherney, Caputo, & Rheinstein, 1987;Kushner, Sher, & Beitman, 1990; Mullan, Gurling, Oppenheim, & Murray, 1986; Bowen,Cipywnyk, D’Arcy, & Keegan, 1984; Powell, Penick, Othmer, Bingham, & Rice, 1982;Weiss & Rosenberg, 1985). In addition, Cox, Norton, Dorward, and Fergusson (1989)found that over 50% of a group of inpatients with alcohol-related diagnoses reported atleast one panic attack in the prior 3 weeks. Over 80% of these patients reported usingalcohol to self-medicate their panic attacks. The above-mentioned studies looked at the incidence of anxiety disorders in individu-als diagnosed with alcohol-related disorders. Another approach involves examining theincidence of alcoholism in patients diagnosed as having an anxiety disorder. Otto, Pol-lack, Sachs, O’Neil, and Rosenbaum (1992) found that 24% of 100 patients with panicdisorder presented with a history of alcohol dependence. Thyer, Parrish, et al. (1986) foundthat 27 out of 156 patients with anxiety disorders, or 17.3%, scored in the “alcoholic”range on the Michigan Alcoholism Screening Test. Bibb and Chambless (1986) also foundthat from 10% to 20% of individuals with agoraphobia were misusing alcohol. The obvious question is this: Which comes first, the anxiety or the substance use prob-lem? Several investigators have looked at this issue, and the evidence emerging suggeststhat a complex relationship exists (Kushner, Sher, & Beitman, 1990; Kushner, Abrams, &Borchardt, 2000). Some retrospective studies indicate that severe anxiety precedes the onsetof drinking or substance misuse in most cases (Chambless et al., 1987; Mullaney & Trippett,1979; Smail et al., 1984). Other studies suggests that alcoholism consistently predates thedevelopment of anxiety (Mullan et al., 1986; Kushner, Sher, & Beitman, 1990). The thoughtful review by Kushner, Sher, and Beitman (1990) of studies done prior tothat time suggests that part of the answer for these discrepant results may be due to differ-ent patterns of alcohol use problems associated with specific anxiety disorders. In the pho-bic disorders (particularly agoraphobia and social phobia), alcohol use problems appearmore likely to follow the development of the disorders and to begin, perhaps, with attemptsto self-medicate the phobic and anxiety symptoms. Disorders without substantial phobiccomponents (such as panic disorder and generalized anxiety disorder [GAD]), on the otherhand, may be more likely to be a consequence of excessive consumption of alcohol. Kushner,Sher, and Beitman (1990) also reported that the incidence of alcohol use problems washigher in patients with phobic disorders than in patients with anxiety disorders withoutsubstantial phobic components. In a later review, Kushner et al. (2000) confirmed thatanxiety disorders and alcohol use disorders can each serve to initiate the other, and thatanxiety contributes to maintenance and relapse of alcohol use disorders. Thus a number of studies have demonstrated elevated incidence of alcohol use prob-lems in patients with anxiety disorders, particularly phobic disorders; and studies have alsosuggested a higher incidence of anxiety and panic in populations suffering from substance
The Experience of Anxiety 15use problems. But the figures differ substantially from study to study, as well as betweencommunity samples assessed though proper epidemiological methods and clinical samples.Some investigators reviewing the literature even found little evidence of an increased riskof true anxiety disorders among individuals with alcohol dependence, concluding that mostprior studies suffered from faulty methodology (Schuckit & Hesselbrock, 1994). Morerecently, however, a major study reported by leading epidemiologists from around the worldhas shed some light on the complexities and contradictions reported above. Swendsen et al. (1998) examined the comorbidity of alcoholism with anxiety anddepressive disorders in four different epidemiological investigations conducted in diversegeographic sites. This study is noteworthy because the four epidemiological investigationsare well known for having used “state-of-the-art” methodology, and because the resultswere relatively consistent in different parts of the world. This consistency affords the firstdefinitive answers to questions on the relationship of substance abuse and dependence toanxiety and depressive disorders noted above. The four epidemiological studies utilized were the ECA study mentioned above, whichsampled over 12,000 community residents in five different geographical areas in the UnitedStates (excluding institutionalized individuals); the National Comorbidity Survey (NCS),a sophisticated probability sample of over 8,000 noninstitutionalized community residents;an epidemiological study of Puerto Rico, which consisted of a stratified sample of the en-tire island, yielding over 1,500 eligible residents; and a cohort study of young adults inZurich, Switzerland, where almost 600 residents of the canton of Zurich were interviewed.Data for the comorbidity of alcoholism with specific anxiety or depressive disorders arepresented in Table 1.2. These data clearly indicate that the probability (the odds ratio) ofpresenting with alcohol abuse or dependence within the past 12 months is two to threetimes greater if individuals have also had a depressive or anxiety disorder at some point intheir lives than if they haven’t. These odds ratios represent statistically significant eleva-tions of risk. Furthermore, these ratios are probably underestimates, since GAD, OCD,and posttraumatic stress disorder (PTSD) were not included in these analyses. Swendsen et al. (1998) also attempted to ascertain the order of onset of alcoholism inrelation to anxiety or depressive disorders. This analysis confirmed the findings of Kushner,Sher, and Beitman (1990), in that no consistent or distinctive pattern was observed con-cerning the onset of alcoholism relative to the nonphobic disorders of major depression,dysthymia, or panic disorder; that is, sometimes alcoholism came first, and at other timesthe anxiety or depressive disorder came first. In contrast, there was strong evidence thatphobias preceded the development of alcohol abuse or dependence. These data are clearlypresented in Table 1.3. Values are provided for the proportion of subjects with the onsetof alcoholism either preceding, concomitant to, or postdating the onset of anxiety or de-pressive disorders. For example, for comorbid depressive disorders and alcoholism in theECA study, 45% of the comorbid sample developed alcoholism first, 10% developed thedisorders at the same time, and 45% developed the depressive disorders first. For phobiasand alcoholism in the ECA study, on the other hand, 13.1% developed the alcoholism first,and 83% developed the phobic disorders first. The investigators conclude that since alcoholism, anxiety disorders, and depressive dis-orders are the most frequent psychological disorders in community samples, the fact thatthey frequently co-occur is of considerable concern to clinicians and public health officials.Furthermore, whereas the order of onset of alcoholism and nonphobic anxiety or depressivedisorders is inconsistent, a strong relationship emerges for alcoholism following the onset ofphobic disorders. Thus panic disorder or GAD could occur either as a consequence of alco-holism, or subsequent to attempts to self-medicate panic and anxiety. For phobic disorders,
16 ANXIETY AND ITS DISORDERSTABLE 1.2. Odds Ratios of Previous 12–Month Alcohol Abuse/Dependence Comorbiditywith Lifetime Depressive and Anxiety Disorders, Accounting for Age, Gender, and Education(Weighted Data) ECA NCS Puerto Rico Zurich OR 95% CI OR 95% CI OR 95% CI OR 95% CIComorbidity of alcohol abuseand/or dependence with . . . Major depression 3.01* 2.89–3.12 2.27* 1.73–2.99 1.25 0.46–3.41 4.77* 1.72–13.22 Dysthymia 1.67* 1.58–1.71 2.45* 1.19–5.06 1.36 0.42–4.40 4.53 0.86–23.95 Any depressive disordera 2.68* 2.63–2.90 2.28* 1.70–3.06 2.88 0.82–10.09 4.48* 1.38–14.53 Agoraphobia 2.67* 2.57–2.78 2.57* 1.55–4.28 2.91* 1.23–6.90 7.84* 2.07–29.73 Simple phobia 1.95* 1.91–1.99 2.91* 2.08–4.06 1.79 0.84–3.86 3.27 0.99–10.82 Social phobia 1.81* 1.01–3.26 2.20* 1.49–3.25 0.87 0.12–10.98 2.96 0.32–27.09 Any phobiab 1.95* 1.92–1.99 2.40* 1.79–3.22 3.46* 1.85–6.48 4.23* 1.05–17.01 Panic disorder 4.06* 3.90–4.22 1.21 0.77–1.90 0.40 0.49–12.68 4.08 0.79–21.18 Any anxiety disorderc 2.10* 2.06–2.14 2.38* 1.77–3.19 3.26* 1.78–5.99 2.32* 1.02–5.29Note. OR, odds ratio; CI, confidence interval. Swendsen et al. (1998). Copyright 1998 by W. B. Saunders Company.Reprinted by permission.aAny major depression or dysthymia.bAny agoraphobia, simple phobia, or social phobia.cAny agoraphobia, simple phobia, social phobia, or panic disorder.*p < .05(two-tailed). on the other hand, the self-medication hypothesis receives strong support and underscores the risk in phobic disorders of developing substance abuse or dependence. Patients with anxiety disorders who also suffer from alcohol abuse or dependence generally present with more severe anxiety (Chambless et al., 1987; Woodruff, Guze, & Clayton, 1972). Periods of abstinence seem to result in a general improvement in fear and anxiety in many patients (e.g., Stockwell, Smail, Hodgson, & Canter, 1984). Thus, con- trary to myth, alcohol does not necessarily reduce anxiety and fear in the long term, and may in fact worsen it (Thyer & Curtis, 1984). Whether anxiety disorders precede substance use disorders or follow them, the alcohol (or drug) use seems to have a deleterious effect on mood, creating a vicious cycle. (Kushner et al., 2000; Kushner, Sher, & Erickson, 1999). This finding is very clearly underscored in the epidemiological data from Swendsen et al. (1998). Table 1.4 presents increases in symptoms of anxiety and depression associated with comorbid alcohol abuse or dependence. Although there was little or no impact of alcohol- ism on the symptoms of phobic disorders, comorbid alcoholism increased other depressive and anxiety symptoms substantially in some cases. Conversely, the presence of any anxiety or depressive disorder was significantly associated with moderate increases in the number of symptoms of alcohol abuse or dependence. Among the anxiety disorders, this effect was most notable for panic disorder, with an increase of 0.48 (ECA) or 0.60 (NCS) in the number of alcohol symptoms over what would be expected from alcohol abuse or dependence alone. A recent prospective study, the first of its kind, has confirmed that the odds of devel- oping a comorbid anxiety disorder in the context of an existing alcohol use disorder (alco- hol use disorder leading to anxiety disorder) or vice versa (anxiety disorder leading to al- cohol use disorder) are 3.5 to 5 times what they would be in the absence of a preexisting disorder (Kushner, Sher, & Erickson, 1999).
The Experience of Anxiety 17TABLE 1.3. Retrospective Estimates for Order of Onset of Alcohol Abuse/Dependence withAnxiety and Depressive Disorders Depressive disordersa Phobiasb PanicAge ofonset ECA NCS Puerto Rico ECA NCS Puerto Rico ECA NCS Puerto RicoA<B 45.0% 54.9% 40.0% 13.1% 19.0% 27.3% 45.2% 62.3% 33.3%A=B 10.0% 10.7% 20.0% 3.2% 4.2% 9.1% 10.2% 5.1% 16.7%A>B 45.0% 34.4% 40.0% 83.0% 76.7% 63.6% 44.5% 32.7% 50.0%Total no. 174 441 15 315 585 33 23 96 6Note. A, alcoholism; B, index disorder. From Swendsen et al. (1998). Copyright 1998 by W. B. Saunders Company.Reprinted by permission.aAny major depression or dysthymia.bAny agoraphobia, simple phobia, or social phobia. Thus anxiety and panic, when self-medicated with alcohol, result in an ever-increasing downward self-destructive spiral—not only from the effects of alcohol (or drug) addic- tion, but also from the exacerbating effects of the drugs on the anxiety and panic. (Kushner et al., 2000). It may be this complication, along with the development of helplessness and depression, that leads to the increased risk of suicide in patients with anxiety (Coryell et al., 1986; Norton et al., 1993). Furthermore, alcohol is not the only substance that has a notable association with anxiety. There is a growing literature on the strong association between the use of cocaine and at least one anxiety disorder: panic disorder (Louie et al., 1996; Bystritsky, Ackerman, & Pasnau, 1991). In one report (Louie et al., 1996), 86 patients reported developing panic attacks after significant usage of cocaine. But the more frightening finding was that full- blown panic disorder continued long after cessation of cocaine use and misuse. Thus panic attacks seem related to both the use of cocaine and the abrupt cessation of its use (almost 20% of patients reported that they were worse immediately after stopping the cocaine), and panic disorder and associated anxiety may linger after cessation of use. The experi- ence of these individuals shows that anxiety and panic are a very heavy price indeed to pay for the momentary pleasure associated with cocaine. Surgery Substance abuse or dependence is often a self-initiated attempt to cope with the unbear- able experience of anxiety. In its most severe form, the paralysis and self-destruction asso- ciated with anxiety have led to equally dramatic attempts on the part of concerned health care practitioners to alleviate this suffering. Among the more desperate efforts is neuro- surgery (sometimes called “psychosurgery,” since the goal is to treat psychological disor- ders) (Jenike et al., 1991; Marks, Birley, & Gelder, 1966; Sachdev, Hay, & Cumming, 1992; Smith & Kiloh, 1980). As with any major surgical procedure, death can be a complica- tion. Causes of death are most often cerebral hemorrhage or edema, although the trend toward more precise and limited lesions has greatly decreased the risk of serious side ef- fects. Neurosurgical procedures as treatments for anxiety disorders, mostly severe OCD, have been reported in the literature on thousands of patients (e.g., Smith & Kiloh, 1980).
18 ANXIETY AND ITS DISORDERS TABLE 1.4. Increase in Disorder Symptoms Associated with Comorbid Alcohol Abuse/Dependence Metric regression Study site Class of symptoms coefficient 95% CI ECA Major depression .96* 0.87–1.05 Dysthymia 1.00* 0.91–1.09 Panic .35* 0.29–0.42 Agoraphobia .01 0.00–0.02 Simple phobia .01 1-01–0.03 Social phobia .00 0.00–0.01 Any phobiaa .04* 0.01–0.07 Any anxietyb .59* 0.49–0.69 NCS Major depression .81* 0.71–0.91 Dysthymia 1.00* 0.90–1.10 Panic .48* 0.36–0.60 Agoraphobia .07* 0.03–0.11 Simple phobia .07* 0.03–0.11 Social phobia .15* 0.09–0.21 Any phobiaa .25* 0.15–0.35 Any anxietyb .58* 0.40–0.76 Note. Global model df (ECA) = 5, 12,708; global model df (NCS) = 6, 8,091. From Swendsen et al. (1998). Copyright 1998 by W. B. Saunders Company. Reprinted by permission. aAny agoraphobia, simple phobia, or social phobia. bAny agoraphobia, simple phobia, social phobia, or panic disorder. *p < .05(two-tailed).Although rare in the United States and the United Kingdom, these therapeutic proceduresare fairly standard in some countries. It is probably safe to conclude that hundreds of thou-sands of patients have consented to this type of surgical treatment, most often after allother treatments have failed. Surgical procedures have become very sophisticated in recent years. For OCD,cingulotomy and orbito-medial lesions are now the most common procedures. Jenikeet al. (1991) and Sachdev et al. (1992) found that between 25% and 40% of patientswith very severe and intractable OCD evidenced some clinically useful improvement (seeChapter 15).PREVALENCE AND COURSEAnxiety kills relatively few people, but many more would welcome death as an alternativeto the paralysis and suffering resulting from anxiety in its severe forms. For decades, stud-ies have shown that millions of individuals each year seek help for what is broadly con-strued as “anxiety” or “nervousness.” Statistics compiled from the offices of front-lineprimary care practitioners startle even the most jaded experts in the area. In an early studyfrom the state of Virginia, investigators surveying the reasons why patients sought out their
The Experience of Anxiety 19local physicians found that hypertension, cuts and bruises, and sore throats ranked rightbehind a general medical checkup as the most common reasons motivating a visit. Closebehind these common maladies was “anxiety,” ranking well ahead of even bad colds orbronchitis (Marsland, Wood, & Mayo, 1976). Another early survey documented that oneof every five primary care patients was taking benzodiazepines (Wells, Goldberg, Brook,& Leake, 1986). This kind of information explains the high utilization of health care andhigh costs of anxiety in primary care settings (e.g., Greenberg et al., 1999; Simon et al.,1995). In fact, one recent study documented that over 50% of the estimated $1,500+ costper individual suffering from anxiety in the United States was attributed to utilization ofprimary care services (Greenberg et al., 1999). Is everyone with a complaint of “anxiety” or “nerves” who seeks out a local physi-cian or health care practitioner really suffering from anxiety? Does everyone who takesminor tranquilizers have a clearly defined anxiety disorder? Most clinicians and investiga-tors would guess that these millions of individuals do not present with clearly identifiableanxiety disorders, but rather with some vague combination of stress, adjustment to diffi-cult family or work situations, or other temporary problems (e.g., difficulty sleeping). Foryears it was impossible actually to ascertain the number of these individuals presentingwith anxiety disorders. Recently, however, epidemiologists have begun to undertake thisarduous task, culminating in two of the most ambitious studies of the prevalence of men-tal disorders ever undertaken. Before these studies are described, it is interesting to put the numbers in context. Oneof the first studies using sophisticated, up-to-date sampling techniques to estimate the dis-tribution of fears and phobias in the general population was undertaken by Agras, Sylvester,and Oliveau (1969). They conducted a probability sample of the household population ofa small city in the United States (Burlington, Vermont), and they interviewed the 325 indi-viduals who made up the sample. From this study, the estimated total prevalence of pho-bias was 7.7%, but only 0.02% presented with phobias severe enough to result in an ab-sence from work or the inability to manage common household tasks. The investigatorsdiagnosed 0.06% of the sample, or 6 out of 1,000, as having agoraphobia. Many moreindividuals, approaching 50% of the population, presented with mild fears of objects orsituations (snakes, heights, storms, etc.). Another important finding from this study was that phobias ran a prolonged course:Most often an individual, once he or she developed a phobia, had that phobia in at least amild form for a lifetime. Figure 1.1 shows the rates of incidence (the beginning of the pho-bia) and prevalence (the presence of the phobia) at various ages within the population.The findings on the prolonged course of phobias were confirmed in a later follow-up study:Agras, Chapin, and Oliveau (1972) found very little improvement in individuals with un-treated phobias followed 5 years after the original study, particularly if they were 20 yearsold or older and if their phobias were more generalized. More recent data supporting thisfinding are presented below (Goisman et al., 1998). In fact, a distinguishing feature ofanxiety disorders and major depressive episodes is that depression tends to remit tempo-rarily whether treated or not—usually in a matter of months, with 9 months the averageduration (Barlow & Durand, 1999). Anxiety disorders, on the other hand, tend to be chronicand to remain present in somewhat less severe form even if successfully treated (Noyes &Clancy, 1976; Noyes, Clancy, Hoenk, & Slymen, 1980; Yonkers, Warshaw, Massion, &Keller, 1996; Roy-Byrne & Cowley, 1995). Two very important developments made more sophisticated and wide-ranging epi-demiological efforts possible. First, diagnostic criteria for anxiety disorders were speci-fied in much more detail, allowing more certain identification of the various anxiety
20 ANXIETY AND ITS DISORDERSFIGURE 1.1. Incidence and prevalence rates for phobias within the general population. From Agras,Sylvester, and Oliveau (1969). Copyright 1969 by Grune & Stratton. Reprinted by permission.disorders. Second, semistructured interviews were devised to insure that all investiga-tors would cover the same essential points when interviewing for the presence of anxiety(or other) disorders (see Chapter 9). These developments permitted a more standardizedand objective approach to the question of the frequency of specific clinical anxiety dis-orders in the population. Data from the ECA study, sponsored by the National Institute of Mental Health anddescribed above, were collected in five different sites around the United States beginningin 1980. The sample population size in each site was approximately 300,000. Approxi-mately 12,000 individuals across sites were interviewed. A semistructured interview thatyielded Diagnostic and Statistical Manual of Mental Disorders, third edition (DSM-III)anxiety disorder diagnoses was employed by lay individuals trained to administer this in-terview. Some of the results for panic and phobic disorders from the first wave of inter-views are presented in Tables 1.5 and 1.6. These tables present results for panic disorderand the three phobic disorders broken down by sex and ethnicity. Agoraphobia, in Table1.5, refers to agoraphobia with and without panic attacks according to the older DSM-IIIcriteria. Table 1.5 provides data for lifetime prevalence only for the phobic disorders. Asone can see from both tables, females in this study evidenced panic and phobic disordersat about twice the rate of males. From the second wave of the ECA study, we also have anestimate of the lifetime prevalence of DSM-III GAD of 4.0% (Blazer, Hughes, George,Schwartz, & Boyer, 1991). The second major epidemiological study, the NCS, referred to above, interviewed arandom sample of over 8,000 noninstitutionalized individuals in the United States. Sincethis was a later study, DSM-III-R diagnoses were utilized. Data indicating the prevalence