Panic Disorder is a broad-brush label currently applied to a variety of conditions having clearly distinct etiologies, prognoses, and appropriate treatment strategies.

Reasonably distinct causes of panic disorder include:

• Phobic sensitization to symptoms of panic or anxiety – "fear of the fear"
• Prolonged stress
• Anxious or obsessive patterns of thinking
• Poor breathing habits, whether rooted in anxiety or other causes
• Emotional or existential turmoil (whether conscious or unconscious), often deriving from conflicting expectations, needs, desires, or external requirements and resulting in a subjective perception of helplessnes or paralysis

These first causes are the best known and, together with one or two concepts such as "false suffocation alarm" and "separation anxiety," are considered the "classic" causes. However, they may in fact explain a minority of cases. Other causes include:

• A hereditary defect in the metabolism of cholecystokinins (CCK), endocrine substances involved with regulating both digestion and stress endorphins. People with this form of panic disorder have likely exhibited escalating gastrointestinal symptomatology, often Gastro-Esophageal Reflux Disease or Irritable Bowel Syndrome, for years or decades before the emergence of panic attacks.
  
  
• A tendency to absorb and react to opioid proteins that are part of the molecular structure of gliadin (a constituent of gluten, a grain protein) and casein (a milk protein). Most people digest these proteins down to the level of amino acids, but certain people lack the enzymes required to do so. People with celiac disease and people who are subject to migraines are especially likely candidates for this mechanism.
  
  
• The neurological and/or psychological effects of adverse reactions to the street drug MDMA / Ecstasy / Xtc (with onset possibly delayed by as much as 4-6 weeks, although it can be nearly immediate) or medical use of selective serotonin reuptake inhibitors (with onset typically occurring, if at all, within 1-7 days). In both cases, the condition sometimes persists after the drug has been eliminated from the body.
  
  
• A yet-to-be-isolated factor that is partially cross-heritable with schizophrenia. Multiple independent studies conclude that a person who has a primary relative with schizophrenia has equal chances of developing either schizophrenia or panic disorder; however, panic disorder in a primary relative is not a predictor for schizophrenia. Additionally, there is a certain comorbidity of panic disorder in schizophrenic patients; and according to a number of sources, patients with panic disorder were frequently misdiagnosed with schizophrenia in the past. People with this form of panic disorder have likely exhibited escalating biological and psychological hyperstimulability for years or decades in advance of the emergence of panic attacks.
  
  
• Irritation of the vagus nerve from mechanical impingement or other factors.
  
  
• Parasympathetic nervous system damage caused by overexposure to neurotoxic compounds, commonly solvents in paints, glues, foams, and other building materials or industrial chemicals.
  
  
• A hereditary factor causing progressive insufficiency of the parasympathetic nervous system. This may be linked to the schizophrenia connection, though there has been little or no study of such a connection as of yet.
  
  
• A congenital or acquired abnormality in the number or sensitivity of adrenergic neuroreceptors throughout the body. This was a prolific and productive area of research until funding was rather abruptly shifted into areas more supportive of certain pharmaceuticals that were recently patented at the time. It is interesting to note that while panic attacks are often described as "false alarms of the fight-or-flight reflex," some studies looking specifically for surges of adrenergic hormones during actual panic attacks have failed to find any such thing. Panic disorder is more complicated than such simplistic characterizations attempting to relate it to universal experience.
  
  
• A yet-to-be-identified factor causing a rare, heritable form of panic disorder in which panic attacks may be extremely severe and resemble seizures closely enough to lead to misdiagnosis. However, such panic attacks never show seizure-like brain activity in an EEG, although they can be seen in a PET scan. Reports regularly occur of this form of panic disorder being misdiagnosed as one or another form of epilepsy. One prominent expert on panic disorder has referred to a panic attack in this variant of panic disorder as "conceptually a seizure that is not in the cortex of the brain," going on to state that the condition is "closely related to anxiety, but is really a different disease."
  
  
• Congenital or acquired abnormalities in brain chemistry possibly distinct from any factors listed above. Different bodies of research implicate different neurotransmitters, principally:
  
  
  • Norepinephrine and secondarily epinephrine
  • GABA
  • Serotonin
    
    While researchers acknowledge the interrelation of these neurotransmitters, their regulation and their effects, it is not clear whether one or more comprise the dog and the other(s) the wagging tail. The advent of Prozac and similar medications targeting serotonin receptors with very high specificity led to a flurry of research and to authoritatively stated claims that serotonin is the master regulator. (This is somewhat analogous to what took place several decades ago with the advent of Valium and similar medications targeting GABA/benzodiazepine receptors with a high degree of specificity.) Commercial and professional vesting in the serotonin doctrine has exerted a profound influence on both research into and treatment of panic disorders. Fortunes and reputations hang in the balance, and so dogmatic positions have been staked out and promoted ferociously – sometimes perhaps unscrupulously. But results of the serotonin doctrine in clinical practice have been mixed at best, and reflect the reality that no single brain-chemistry-based model describes all or even a preponderance of cases (a situation that should be obvious from the variety of etiologies described above). And when one considers the two-layer model of panic attacks and the medical comorbidities discussed in other sections of this document, it looks less certain that the truly fundamental causes of all types of panic disorders necessarily lie anywhere within the brain.
    
    
• A postulated "chromosome 13 syndrome" identified by carefully blinded genetic and statistical analysis. According to the researchers who advocate this hypothesis, a hereditary abnormality in a certain area of chromosome 13 expresses itself in a variety of forms including panic disorder, kidney or bladder problems, gut problems, serious headaches, thyroid problems (usually hypothyroid), mitral valve prolapse, and/or schizophrenia. Because the abnormality manifests itself in such a variety of ways, the existence of a "familial tendency" is often masked because patients and doctors would not associate, for instance, bladder problems in one family member and mitral valve prolapse in another, even though the genetic evidence supports the potential for a common causality. Although the researchers appear to be the first to postulate a "chromosome 13 syndrome" as they describe it, their conclusions appear to mesh well with other research and clinical observations on panic disorder and its associated conditions (for which, see below).
  
  
• Decreased integrity of the blood-brain barrier in conjunction with the absence of certain protease enzymes. Recent "smoking gun" evidence confirms older and heretofore controversial behavioral research on the effects of diet in some cases of schizophrenia and autism. Researchers demonstrated in 1999 that milk and grain proteins which circulate at hundred-fold levels in the blood of many schizophrenic and autistic individuals also concentrate in 32 areas of brain tissue, although it was previously thought impossible for such complex food proteins to cross the blood-brain barrier. The same research implicates lack of certain digestive enzymes and the presence of immune system responses to the proteins in question. Many of those 32 affected brain areas are directly involved in the symptoms of schizophrenia and autism. Increased permeability of the BBB has been suspected in earlier panic disorder research, and it is known that one of the effects of tricyclic antidepressants (which are sometimes effective against panic disorder) is to enhance function of the barrier. Proof that panic disorder can result through decreased function of the blood-brain barrier is still apparently lacking, but discovery of these proteins in brain tissue is certain to lead to a search for other pathogenic dietary proteins (and possibly inhaled air pollutants as well). This would seem likely to lead to solid neurological explanations for connections that are already informally recognized between diet and panic disorders.
  
  

Each of the causes of panic disorder listed above is supported by a body of published research, except as noted in the last point, where the relevance to panic disorder is more speculative and is not suggested by the researchers but is suggested by the well-documented statistical association of panic disorder with schizophrenic families. Ongoing advancement in the scientific understanding of panic disorder may reveal etiological overlap among these factors beyond the overlap already suggested or it may lead to further differentiation.

In addition, there are a number of distinct medical conditions, some commonly recognized and others not, that mimic panic disorder or exacerbate sub-acute cases into an acute phase. Beyond the universally recognized but very rare pheochromocytoma, some of the more commonly encountered conditions are:

• Hypothyroidism. Screening for TSH levels is widely regarded among first-line clinicians as sufficient screening for thyroid abnormalities; however, such limited screening consistently fails to identify a number of hypothyroid conditions that lead to panic attacks and other serious and sometimes life-threatening pathology. Once the condition has advanced to the stage of panic disorder, correcting the underlying imbalance may not be sufficient to bring about remission, as damage to the nervous system can already have occurred at this point.
  
  
• Celiac Disease and other autoimmune or inflammatory conditions associated with intolerance to dietary gluten. In the late 1990s, public health authorities in the United States became aware of preliminary research indicating that the incidence of celiac disease in this country is probably similar to that across Europe, namely somewhere between 1:150 and 1:250. This came as a surprise, because the previous estimates ranged from 1:3500 to 1:9000. This means that for every case that is diagnosed, there are between 14 and 60 cases that are misdiagnosed as something like Irritable Bowel Syndrome or are not diagnosed at all. Public health efforts are being made to re-educate doctors and especially gastroenterologists in the true picture of celiac disease, but progress is frustratingly slow. Because there is no drug treatment at present (except for extremely advanced cases), there is no incentive for pharmaceutical companies, from whom physicians receive most of their continuing education, to emphasize such information. It is in the pharmaceutical companies' interests to continue the treatment of refractory bowel disease with psychotropic medications. But panic attacks and the associated syndrome of panic disorder are in fact a frequent feature of undiagnosed celiac disease and other autoimmune and inflammatory conditions caused by the false alarm reaction of the immune system, primarily in the form of IgG antibodies. According to the European research that American researchers are only starting to discover and replicate, anyone with significant levels of IgG antibodies to gliadin is at a greatly elevated risk of autoimmune disease of some form, whether thyroid, pancreatic, arthritic, or celiac, to name the most obvious. The correct tests for diagnosing celiac disease in particular look for a combination of anti-gliadin, anti-endomysium, and anti-jejunum antibodies. There is also a trend toward the use of a test for anti-transglutaminase antibodies, a test that is claimed to be more accurate all by itself than the current standard of endoscopy (which has an unacceptably high rate of false negatives, largely owing to the sometimes localized nature of the characteristic lesions). However, the presence of IgG anti-gliadin antibodies is enough in itself, even in the absence of gastrointestinal complaint, to predict that a patient is absorbing constituent opioid peptides at least from gluten and often from casein as well.
  
  
• Depression. Panic attacks and a syndrome diagnosable as panic disorder may be secondary to depression. However, in panic disorder patients it is more common to find that depression is secondary to panic disorder and resolves when the panic disorder is brought under control even without the use of antidepressant medications. It should be no surprise that panic disorder arises more often from disorders of arousal than from disorders of depression.
  
  
• Aldosteronism. There are a variety of reasons that the adrenal glands may produce an excess of the hormone aldosterone. The best-known role of aldosterone is to regulate sodium/potassium balance in the body. When the ratio of aldosterone to renin (a kidney-regulating hormone) is too high, the body has trouble holding on to potassium and may develop an alkaline state that feels much like hyperventilation but cannot be corrected by changes in breathing. Aldosterone receptors are located throughout the body, not only in the kidneys (which regulate the balance of sodium and potassium). In fact, the largest concentration of aldosterone receptors is in the brain. Elevated aldosterone levels suppress parasympathetic function (the calming function) in the nervous system. The result is an entire cascade of potential problems in areas like digestion (and gut inflammation), insulin metabolism, bladder function, heart rhythm, and disposal of stress and startle hormones. It was once thought that all people with aldosteronism showed deficient potassium in routine blood tests, but it is now established that aldosteronism cannot be ruled out without comparing plasma aldosterone concentration to plasma renin activity. It was once thought that aldosteronism was rare, representing as little as 0.5% of the hypertensive population, but later research suggests that as much as 10-15% of hypertension is secondary to aldosteronism. The cut-off levels once used to evaluate aldosterone-to-renin levels were once very high and correlated with certainty of adrenal tumors. Today considerably lower levels are applied, levels that correlate with symptoms that respond to treatment. Experts today concur that any ratio over 14 provides at least 96% certainty of aldosteronism and the certainty rises steeply as the ratio climbs above 14. Furthermore, people with sensitive nervous systems may experience symptoms at lower PAC-to-PRA ratios. It is rare to find a doctor, even an endocrinologist or psychiatrist, who appreciates the extent to which aldosteronism can destabilize the nervous system by suppressing parasympathetic function.
  
  
• Other Forms Of Endocrine Imbalance. Although endocrine problems can be some of the more complex conditions tied to panic disorder, some cases can be relatively simple. Deficiencies in progesterone, or more likely unfavorable progesterone/estrogen ratios, are often implicated in panic disorder. It is well known that panic disorder often goes into remission during pregnancy, with the greatly elevated levels of progesterone experienced then. It is also not uncommon for women to experience postpartum panic attacks as their progesterone levels decline precipitously. (Such panic attacks do not, however, as a rule constitute panic disorder.) Abnormal testosterone levels can play a role in panic disorder, as can excessive levels of insulin in the presence of normal blood sugar levels. It is also common to see abnormal levels of cortisol in panic disorder patients, either elevation or depletion. Beyond these examples, some of the most severe (and fortunately very rare) instances of panic disorders have been found to have connections that are endocrinological in the traditional sense. These cases have been very difficult to treat and have left endocrinologists and neurologists at teaching hospitals shaking their heads and resorting to improvisation to save their patients' lives.
  
  
• Gut pathology other than celiac disease. In the case of panic disorder there is often a chicken-and-egg question surrounding any gut pathology. But the gastrointestinal tract is the body's largest client of the parasympathetic nervous system, and any derangement that stresses the GI tract has the potential to derange the parasympathetic nervous system, with a consequent reduction in its autonomically balancing effects.
  
  
• Mitral valve prolapse, an often overlooked (and generally benign) condition in which a valve in the heart fails to close completely, allowing some turbulence and localized backflow of the blood. The problem is often familial and often closely linked to symptoms resembling those of panic disorder but can easily be differentiated by ultrasonic examination.

Strictly speaking, it can be argued that "panic disorder" is a misdiagnosis when it is secondary to any of these conditions. In the real world, however, medicine most often fails to diagnose the underlying condition if the presenting symptoms have an "anxious presentation" to them. As a result, the "panic disorder population" – on which all research is conducted, from which the understanding of panic disorder derives, on which treatments are tested, and to which the resulting diagnostic and therapeutic protocols are applied – is exactly this mixed bag of patients and causes. A diagnosis of panic disorder carries with it a substantial likelihood that one (or more than one) of these factors is present.

Studies over long periods have established that thought patterns once thought to be causative in panic disorder, such as hypervigilance, catastrophic thinking or catastrophic interpretation of sensations, among others, emerge in many patients as a result of the condition and cannot in fact be considered preexisting causative factors in those patients – that, for instance, the tendency toward catastrophic interpretation of relatively normal physical sensations typically emerges (if at all) as a gradual consequence of the "normal" interpretation of catastrophic physical sensations lying outside non-afflicted individuals' realm of experience. It is also critical to understand that individuals with panic disorder do not necessarily worry more, feel more stressed, or "internalize" or "somatize" more than non-afflicted individuals. They can certainly develop such traits as a result of the condition (or possess them from birth), but the distinguishing factor is that the neurochemistry of an individual with panic disorder reacts differently from that of a non-afflicted individual even to low or normal levels of stress – or to other, idiopathic factors not necessarily related to situational stress or worry. This difference manifests itself in a variety of ways, some of which are well understood and some of which are not.

Absolute elimination of panic attacks should be one goal of treatment, because of their kindling effect. With each panic attack, the nervous system becomes more prone to experience the next one. This is not only a psychological process, but a neurological one as well.

Absolute elimination of panic attacks is not a sufficient goal of treatment, much as elimination of sugar from the urine of a diabetic is not a sufficient goal of treatment for diabetes. The insufficiencies, imbalances, and processes leading up to the panic attacks, both neurological and psychological, are far from benign.

Because of the variety of causes and processes at work in panic disorder, there are few generalizations one can make that would apply to as much as 75% of the panic population. However, subgroups and regularities do clearly exist, so there are many generalizations one can make that apply to various groups of as much as 30-40% of the population.

The American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV™) describes a panic attack as follows:

The essential feature of a Panic Attack is a discrete period of intense fear or discomfort that is accompanied by at least 4 of 13 somatic or cognitive symptoms. The attack has a sudden onset and builds to a peak rapidly (usually in 10 minutes or less) and is often accompanied by a sense of imminent danger or impending doom and an urge to escape. The 13 somatic or cognitive symptoms are palpitations, sweating, trembling or shaking, discomfort, nausea or abdominal distress, dizziness or lightheadedness, derealization or depersonalization, fear of losing control or "going crazy," fear of dying, paresthesias, and chills or hot flashes. Attacks that meet all other criteria but that have fewer than 4 somatic or cognitive symptoms are referred to as limited-symptom attacks.

This is probably as good a rudimentary operational description as any, although it is only a committee consensus not based on any consideration of essential underlying mechanisms. (The idea of peaking within 10 minutes is now generally agreed in the panic disorder community to apply to a minority of cases. Full status panicus can persist for hours or even days. Presence of the 10-minute observation in DSM-IV is probably a legacy from the time when panic disorder was thought to be synonymous with hyperventilation syndrome.) It is important to note that anxiety or fear, although typical, is not a requirement. Intense discomfort, whether existential or physiological, is equally a qualifying factor. In fact, the terms "panic disorder" and "panic attack" may do as much to obscure the true face of the disorder from clinicians and patients as they do to make it understandable to the general public.

Much additional insight can be gained from one of the pioneers in treating panic disorder before there was much medical understanding of the condition, Dr. Claire Weekes. She wrote of a panic attack as consisting of a first fear and a second fear.

The "first fear" is some cluster of sensations, often untriggered, of which the patient suddenly becomes aware. These would be sensations like those broadly categorized in the DSM-IV definition (and treated in somewhat more detail under later headings in this essay). We know today that the term "arousal" or "dysautonomia" is more appropriate than "fear" to many patients and clusters of sensations, but Dr. Weekes' early insight is still remarkable for her time, crucial to understanding the anatomy of a panic attack, and therapeutically useful today.

The "second fear" is a reaction to the appearance and sensations of the "first fear." Competing descriptions have emerged for why this second fear comes about – premonition of death, fear of suffocation, resonance with primal separation trauma, catastrophic thinking, etc. – but basically, Weekes' premise is that the second fear is fear of the first fear. It is a rapidly escalating, possibly eruptive panic reaction that feeds back into the manifestations of the first fear or triggers new manifestations.

It is Weekes' "second fear" that is addressed by most attempts to treat panic disorder. Patients are taught not to fear the primary sensations of the episode. This is an extremely useful concept in Cognitive-Behavioral Therapy for panic attacks. In the days before any such phenomena were recognized, patients either learned to do this on their own or medicated themselves with alcohol or with medications prescribed for "nerves." (Or in the worst cases, they were committed to psychiatric hospitals.) The aim of such CBT is to learn and internalize the idea that the primary manifestations are benign and to learn not to compound them with a panic reaction. Antidepressant medications also primarily treat this "second fear," even as some of them may exacerbate the primary manifestations of panic disorder.

Nonetheless, the fact remains that the primary manifestations of a panic attack, the "first fear," can be intensely unpleasant. Even without any "second fear," the primary manifestations of a panic attack can have a distinctly paroxysmal quality of their own, akin to vomiting or a slow-motion sneeze or a slow-motion shudder or an orgasm. They have differing causes in different cases. Various theories compete to provide an explanation. Some neurologists hypothesize neuroelectrical chaos. Some make an analogy to a sort of epilepsy outside the brain, while others dispute the appropriateness or desirability of such an analogy. Others find deficiencies of parasympathethic neurotransmitters. In some patients, blood oxygen saturation drops by 25% or more, sometimes in response to circulatory disturbances and other times in response to underbreathing. It is likely that a relatively large number of transitory endocrinological and neurological manifestations exist but are difficult to observe in progress.

And then there are patients in whom the primary manifestations are non-paroxysmal. Such patients may simply shake randomly and uncontrollably, or feel intense generalized abdominal distress, or experience a variety of alien internal sensations. These sensations can be frustratingly difficult to describe, since they often lie outside of shared experience, the language lacks any directly descriptive terms for them, and the non-afflicted can only try to interpret such descriptions in terms of personal experience. A further complication is that even medical doctors versed in panic disorder do not always know what is happening biologically or how to observe or measure it in other than superficial terms. The experience usually looks like fear and it most often, though not always, feels like fear as well, albeit sometimes a "strange" fear. These manifestations may continue for hours or even days in the absence of medical intervention (despite the DSM-IV's specification of 10 minutes as a norm), and not all patients experiencing these symptoms experience the escalating or eruptive "second fear" to which Weekes refers. It is possible to learn not to compound one's reaction when blood oxygen saturation falls substantially, or when a drop in the availability of the neurotransmitter acetylcholine below a certain threshold causes muscles to go into random spasms, or when any of a host of other manifestations appear, but absence of a compounding "second fear" by no means implies that one is well or feels well. One might well say of some cases – and such cases are not rare – that over time the panic disappears but the disorder remains. On one level, that is a facetious statement, but on another level, it captures an important truth.

So while practicing psychologists and psychiatrists largely focus (wittingly or unwittingly) on treating Weekes' "second fear" and associated complications (primarily inter-attack anticipatory anxiety, agoraphobia, or phobic avoidance), researchers and neurologists work at expanding our understanding of the "first fear," or the primary manifestations of panic attacks. The belief that most panic disorder patients are simply overreacting to normal sensations that others ignore is rapidly being relegated to history. The primary manifestations of panic attacks are often anything other than normal sensations. Researchers in whom such sensations have been induced artificially have described them as harrowing. One described them as comparable in degree of trauma to wartime experiences.

The conditions commonly called anxiety disorders – Generalized Anxiety Disorder, Panic Disorder, Social Anxiety Disorder, Obsessive-Compulsive Disorder, and other less commonly diagnosed conditions – are all essentially disorders of hyperarousal. In the case of panic disorder, particularly of the familial variety, the hyperarousal tends to be essential (having no single, clearly identifiable cause, much like essential high blood pressure) and, at least in the early years, autonomic (expressing itself primarily or initially in physical symptoms not under conscious control).

There has been a tendency in the conceptualization of these disorders (at least the first three) to think in terms of hyperarousal specifically of the fear, alarm, and watchfulness "circuits" of the brain, but in truth the hyperarousal tends to be more generalized in nature. It may even lack an intuitively "anxious" presentation. Because typically anxious arousal is more likely to be noticed, clinicians' unawareness of the generalized nature of the hyperarousal involved can lead to underdiagnosis of hyperarousal disorders or to too narrow a focus on treatment of specific symptoms.

Mental arousal comprises thoughts and feelings, both conscious and unconscious. It may express itself in many ways, some of them subjectively negative (such as worry or non-specific agitation), some of them subjectively positive (such as anticipation), some of them constructive (such as productive focus on a project), some of them destructive (such as constant worry about improbable "what-if" scenarios). Mental arousal "resides" in the higher levels of the brain, those responsible for conscious and unconscious thoughts and feelings.

Autonomic arousal consists of many bodily functions that are not under volitional control, such as heartbeat, sweating, digestion, and others; and some bodily functions that are partially under volitional control, such as breathing and excretion, so that they self-regulate in normal circumstances but are subject to conscious or unconscious intervention from higher levels of the brain. Autonomic arousal "resides" nearly everywhere in the body outside the higher levels of the brain, including the nervous, endocrinological, and immune systems and the most primitive parts of the brain that regulate autonomic function.

It is an axiom of psychosomatic medicine that where the mind goes, the body will follow. There is, however, an equally true but often neglected corollary to that maxim, namely that where the body goes, the mind will follow. Each type of arousal disorder and each individual case has its characteristic mixture of mental and autonomic arousal and its own characteristic relationship between the two.

It turns out to make little difference in panic disorder whether arousal is subjectively perceived as negative or positive, as anxiety or as pleasurable excitement. It is the fact of arousal in itself that primes the body to experience panic attacks and then triggers the attacks. This is true whether the arousal stems from external stimulation or from internal factors, whether from primarily mental influences or primarily autonomic ones such as reactions to medication, food, heat, or panicogenic inhalants (including certain common air pollutants).

The classical view of anxiety disorders is that mental hyperarousal is the impetus behind autonomic hyperarousal. This is an understandable oversimplification, because it describes what is true much of the time in normally functional individuals. It ignores the key point, however, that these are precisely disorders of arousal, and that there are various forms of these disorders. An arousal disorder may be driven by intrinsic hyperarousal of the mind, by intrinsic hyperarousal of the autonomic nervous system, or by an aberration in the way either one influences the other. And whatever the primary impetus of the disorder, secondary effects nearly always emerge to complicate the condition and confuse its characterization – effects seen in the mental and autonomic domains and in the way that the two domains influence each other.

Yes, the mind can arouse the body and it can calm the body. It can even make the body ill, as one sees in the traditional view of stress and psychosomatic medicine; and it can heal the body, as one sees in the response of such conditions as cancer to the influence of psychotherapy – although thankfully, one does not see anyone advocating treatment of the mind as the primary care for cancer patients. (One hopes that the average treatment of panic disorder will some day be so enlightened…)

However, the reverse is equally true: the body can arouse the mind and it can calm the mind; and it can even make the mind ill or heal the mind. One sees one obvious example in the relaxing (or in some cases invigorating) effects of massage – indeed the ability of massage therapy to influence the course of mental and even physical illness. Another example is found in the mental illnesses – depression, obsession, anxiety – that frequently develop as secondary consequences of chronic organic illness. Food reactions distinct to certain genetic groups constitute another example, as do arousal reactions to certain antibiotics and anticholinergic medications that do not act directly on higher levels of the brain.

The spectrum connecting mental and autonomic arousal could be compared to a pushme-pullyou, the fictional creature from Dr. Dolittle with two front ends and two brains. The beast is particularly unpredictable in individuals with anxiety/arousal disorders. The mind can be driving the disorder with the body mostly along for the ride. Or the body can be driving the disorder, with the mind doing its best to resist. Treatment of the mind can accomplish surprising effects on biology, but this is compensation, and it does have its limits. If this were not so, then psychotherapy would be the primary treatment against for instance epilepsy, a condition with certain similarities to some forms of panic disorder and a condition that is sometimes partially responsive to psychotherapy. So a good understanding of the mind-body connection in panic disorder generally and in any specific patient is essential to intelligent diagnosis and effective treatment. Notwithstanding the mind-body overlap, the essence of each case is usually to be found in some intrinsic factor on one side or the other.

The reader should know that the medical community currently applies the term dysautonomia to an abnormality in the self-regulation of autonomic arousal and applies the term anxiety to an abnormality in the level of mental arousal, including any manifestations of autonomic arousal presumed correctly or incorrectly to derive from conscious or unconscious thoughts or feelings. This conceptual bias tends to hobble treatment, but it is important for the reader to understand the doctor's likely perspective at the outset of the relationship. Most of the clinical progress currently taking place in the medical treatment of panic disorder occurs through the process of patients exposing their doctors to the body of research relevant to their own variant of panic disorder.

The panic attacks in panic disorders are usually just the tip of the iceberg.

Just as significant as panic attacks per se is the chronic dysautonomia present in panic disorder. A number of expressions of dysautonomia are often present for extended periods in panic disorder patients, even in the absence of frank panic attacks. Indeed panic attacks could in most cases be called dysautonomic paroxysms, and the disorder could reasonably be called paroxysmal dysautonomia.

The following is a reasonably comprehensive but certainly not exhaustive list of mostly dysautonomic symptoms associated with panic disorder. (Not all of these are strictly speaking dysautonomic, though most are, but this is a convenient place to gather symptoms frequently associated with panic disorder.) Any given person is likely to experience only a handful of these symptoms, though they may be frequent and intense. The point of such an extensive list is to provide reassurance to someone diagnosed with panic disorder who is experiencing any of these very diverse symptoms that they can indeed be merely symptomatic of panic disorder and not necessarily a sign that some other medical or biological condition is present.

• Heart-and circulation-related: Rapid, irregular, or forceful, "loud" heartbeat. Sensation of skipped heartbeats. Chest pain. Low or high blood pressure, often with remarkably wide swings. Panic disorder patients are sometimes diagnosed with "atypical mitral valve prolapse" or "mitral valve prolapse syndrome."
  
  
• Breathing-related: Lowered oxygen level, elevated or lowered carbon dioxide level, under- and over-breathing, sensations of slow suffocation, sensations of inability to take a deep or satisfying breath or to yawn, sensations of abdominal constriction of lungs, often worse when standing or floating chest-deep in water. It is important to note that exposure to high ambient levels of carbon dioxide, as in closed and crowded spaces, or when rebreathing through a paper bag, is a fairly reliable way to provoke a panic attack in someone with panic disorder. Very high humidity can have a similar effect. People with metabolic alkalosis from aldosteronism will feel like they are hyperventilating even when they are not. They are likely to instinctively underbreathe, accumulating acidic carbon dioxide to offset the alkalinity, but subjecting themselves to chronic under-oxygenation in the process
  
  
• Muscle-, skin-, and nerve-related: Jerking, trembling, twitching, spasm, tremor, rigidity, weakness, numbness, tingling, itching, feverishness, chills, hot flashes, "goose bumps," pins-and-needles, isolated "pin jabs," "jelly legs," "legs of lead," excessive or insufficient sweating, intolerance to heat, inability to stretch satisfyingly, flushing, a sudden whole-body, tactile and auditory "whoosh" sensation (either without external trigger or in reaction to minute sounds or other triggers). The whoosh often sounds like having a wave crash around your head or diving into a swimming pool. Hypersensitive conventional startle response, hypersensitivity to even mild stimulants, hyper-reactivity to many medications (not only stimulant medications). Some of these symptoms can be related to over-breathing, or just as commonly to under-breathing – and long-term panic disorder patients can usually tell the difference – but they are often due to other, more inherent factors.
  
  Additionally, panic disorder patients often report a markedly reduced tolerance to exertion, which may cause sensations more akin to a sudden fright than to the tiredness they would have experienced before developing panic disorder. Recovery from these sensations takes much longer than normal recovery from exertion. There tends to be a similar prolongation of the recovery phase from actual sudden frights or confrontations. Where a normal person who experienced a near-accident in morning traffic might experience "jelly legs" and overall shakiness but recover over the space of perhaps 15 minutes to an hour, a person with panic disorder would quite likely still be experiencing those sensations at bedtime. Impaired recovery from such adrenergic surges is one of the most characteristic symptoms patients report.
  
  
• Digestion-related: Tightness in the throat, lump in the throat, sore throat, sensation of foreign object in throat, difficulty swallowing, reflux of acid or bile, nausea, vomiting, reflux asthma, reflux sinusitis, irregularity, spasms, burning or aching pain, bleeding, sensitivity to specific foods (triggering gut symptoms or symptoms from any of the other groups listed here), constipation, diarrhea, incontinence, urinary frequency, spastic bladder, excess secretion of bile, insufficient secretion of pancreatic enzymes.
  
  
• Endocrine-related: Abnormal fluctuations of epinephrine, norepinephrine, insulin, glucagon, renin, aldosterone, cholecystokinin, endorphins, and other endocrine substances. These are normally not noticed except through their expression in other symptoms, but direct attention to endocrine balances is sometimes a factor in the successful management of panic disorder.
  
  
• Emotional-behavioral expressions: (Not all patients with panic disorder experience its expression in the emotional and behavioral domains, and some who do so in early years of the disorder become emotionally desensitized to the dysautonomic and other symptoms without significant reduction in the primary manifestations or their paroxysms. This is more the exception than the rule, however, and many patients with uncontrolled panic disorder enter a degenerative spiral.) Intense fear, anticipation of impending death, anticipation of imminent fainting or blackout (which does sometimes happen, contrary to instruction that some patients receive, but happens rarely), fear of fleeing, fear of "going crazy," sensations of being outside one's body, sensations of unreality, volcanic anger, frequent and escalating anxiety about the next episode, avoidance of situations or places where episodes have occurred in the past, hypochondria, hypervigilance for sensations or the tendency to interpret them catastrophically, low threshold for frustration, low tolerance for forced immobility (as when standing in lines or sitting in traffic). These are all reactions to dysautonomic symptoms rather than dysautonomic manifestations in themselves, but they normally feed back into the physiological experience. Agoraphobia sometimes develops, with patients retreating to smaller and smaller "safe zones" close to home or sometimes within limited areas of the home itself. Spatial/geographic self-limiting sometimes occurs with pre-emergent panic disorder, becoming one of the first behavioral signs of what is to come, as a patient who has not yet experienced panic attacks grows more and more reluctant to leave areas of familiarity and comfort.

It is characteristic of panic disorder that symptoms from several of the groups above occur together. (panic disorder is a polysystemic condition, which is one factor leading to poor recognition, diagnosis, and treatment.) It is also normal for some varying level of these symptoms to be present much or most of the time for uncontrolled patients, but for the symptoms periodically to escalate to the point of paroxysm followed by relief of varying duration. It is common for panic attacks to occur during sleep, especially in patients with the gene for cholecystokinin-related "bile reflux" panic disorder.

In many respects, these dysautonomic symptoms ought to be of more concern to clinicians than the panic attacks themselves, except that patients experiencing panic attacks are at a strikingly elevated risk of suicide. Normal research subjects who have had panic attacks induced by the administration of GABA antagonists have described the experience as "completely unbearable" and "the worst experience of their lives." (It is conceptually interesting that although synthetic panic attacks can be created with some degree of success under laboratory conditions using serotonin depleters, serotonin reuptake inhibitors, cholecystokinin, sodium lactate, and norepinephrine, no class of substances but perhaps cholecystokinin comes anywhere close to the panicogenic effect of a GABA antagonist. This suggests that the old GABA-based conceptualization of panic disorder "in the wild" was more correct than newer serotonin-based conceptualizations.)

In a significant number of patients, dysautonomic symptoms and panic attacks emerge not during periods of long-term or short-term stress, but after the cessation of such stress or during its period of abatement. This is probably true in particular of the patients with cholecystokinin abnormalities.

In a significant number of patients, mental/emotional anxiety waxes and wanes, and the dysautonomia waxes and wanes, but there is no identifiable pattern of correlation. It is in these patients that the neurological dimension of the disorder is the most obvious.

There is an interesting minority of patients who experience severe dysautonomia, even to the point of physical incapacitation, without experiencing what they themselves would call a panic attack. This is not too difficult for many sufferers of panic disorder to understand, particularly after therapy or years of experience with the condition have desensitized (to varying degrees) the fear response that can be aroused by the dysautonomia. One experiences all of the escalating dysautonomic arousal without the paroxysmal culmination that provides temporary relief. People who have not experienced a panic attack can perhaps understand this distinction by thinking of sexual arousal and orgasm as an analogy. It is possible to be highly aroused without experiencing orgasm. Similarly, these non-paroxysmal dysautonomic states can be very intense – indeed they can be more intense than the states that would induce paroxysmal panic in other panic disorder patients or in normal individuals.

Current thinking distinguishes between Generalized Anxiety Disorder (GAD) and panic disorder largely (though not entirely) on the basis of absence or presence of panic attacks. But in some respects, these conditions appear much of the time to be differing degrees of the same biological disorder, and if it is useful to distinguish at all, it might be more diagnostically and therapeutically useful to apply a distinction at right angles to the current one, recognizing one disorder that is primarily somatic/dysautonomic and another that is primarily cognitive/emotional. Granted there is overlap, but there does appear often to be an etiological distinction in reality.

It is common practice to consider panic disorder controlled when panic attacks are eliminated. However, the goals of treatment should actually include elimination of most or all of the associated dysautonomia, as the dysautonomia is symptomatic of uncontrolled underlying pathology and in some cases exacerbates the underlying pathology (just as uncontrolled panic attacks escalate through neurological as well as psychological mechanisms).

Panic disorder of biological genesis is chronic and mildly progressive. Although some clinicians claim encouraging "cure" rates, usually in carefully selected groups, studies have shown that when one looks at patients over periods of time four years or longer, a preponderance of patients experience relapse in the absence of ongoing pharmaceutical treatment. This is not always evident to treating clinicians (who have been known to claim cure rates as high as a thoroughly suspicious 100%), because patients frequently change doctors if they perceive that their treatment was unsympathetic or ineffective; and relapse is readily perceived as a sign that treatment was ineffective if the treating physician has created unrealistic expectations of a permanent cure or has exerted subtle pressure to recover or to "tough out" minor episodes.

It is often stated that panic attacks do not lead to heart attacks. This is true as far as it goes; however, sufferers of uncontrolled panic disorder are at a quadrupled risk for eventually experiencing a serious cardiac event. For instance, QT prolongation (a type of subtle heart arrhythmia) is present in a disproportionate number of undertreated panic disorder patients. In fact, a variety of arrhythmias are observed with disproportionate frequency in panic disorder patients, though many of them are only observable during an episode or during the lead-in to an episode. The paroxysm of the attack itself may serve to correct the arrhythmia.

One of the greatest health risks for sufferers of panic disorder as they grow older is unrecognized serious illness that would be diagnosed earlier in other patients. If it is true that a certain proportion of panic disorder patients tend toward hypervigilance and hypochondria, particularly in the years just prior to and following correct diagnosis with panic disorder, yet it is also documented that primary care physicians and many specialists tend to dismiss the complaints of patients bearing a diagnosis of panic disorder (or generalized anxiety disorder) without sufficient foundation for doing so. Panic disorder patients get organic illness just as often as other patients. They just don't get it diagnosed as often. In fact, it has been clearly shown that patients with anxiety disorders go longer before diagnosis of serious conditions than do patients without these stigmatizing diagnoses, and physicians have cause to be more concerned with the potential for medical error and liability.

Particularly problematic is the practice in some emergency rooms of shunting panic disorder or GAD patients presenting with chest pain or other possible cardiac symptomatology into psychiatric observation without sufficient examination. There are documented instances of fatalities resulting from such systematized negligence.

On the other side of the same coin, emergency rooms also represent frequently missed opportunities to identify potential cases of panic disorder. Many ER doctors are poorly trained to recognize panic disorder, particularly in its pre-emergent guise of non-paroxysmal dysautonomia. This is not surprising given the diagnostic criteria in use today. Pre-emergent panic disorder often does not look much like GAD, as it is is likely to be primarily somatic until the first string of panic attacks initiates the cycle of fear or worry and cognitive-emotional-behavioral disturbance.

Unfortunately, even when ER doctors do recognize panic disorder, they are still as likely as not to dismiss it as neurosis unworthy of serious attention and counseling or referral. Patients are still likely to be told to "breathe into a paper bag next time," although this is in fact more likely to trigger than to stave off a panic attack for a sufferer of biologically rooted panic disorder.

In fact, panic disorder patients are often told (not asked) that they are claustrophobic, or anxious travelers, when in fact their bodies are reacting in predictable and well-understood ways to the elevated levels of CO2 present in poorly ventilated spaces full of breathing bodies. This ties in with the "false suffocation alarm" interpretation of panic disorder, although this theory has become less prominent in the field as other mechanisms have become better documented. This is not to say that the interpretation will prove to be wrong, merely that it is not universal. And in point of fact, it has not as yet been proven to be anything; hypotheses and assertions relating to panic disorder must be subjected to the same standard of proof as other, non-psychology-related, statements of medical doctrine. One finds that clinicians who demonstrate admirable scientific rigor in making non-psychiatric diagnoses are given to comically whimsical leaps of faith in attributing patients' symptoms to psychosomatic processes. (Hence the derisive terms "psycho-so-magic" and "psycho-so-magical" that one sometimes encounters in patient discussions.)

Note: There is ample research around the "false suffocation alarm" to indicate that elevated CO2 levels do in fact provoke hyperarousal and panic attacks in a subset of the panic disorder population. At issue is the histrionic labeling of this reaction on the assumption that all individuals ought to experience identical toxic reactions to CO2 at identical concentrations. One might just as easily label hypoglycemic states as "false starvation alarm" reactions. It is not the alarm per se that is false, it is merely its characterization as one of life-threatening "suffocation" or "starvation" that is false – and that is not the patient's error of interpretation, as the term implies, but the diagnostician's. (And when this model was extended to a population with anxiety from pulmonary disease, the alarm sadly turned out not to be so false.)

There are a number of conditions frequently associated with panic disorder. Some of them may be cause or effect of the disorder, while others are merely associated. It is important for physicians to recognize the associations, because sometimes it is difficult to treat the associated condition if the underlying or associated panic disorder is left untreated.

• GERD, IBS. Probably the conditions most universally present in panic disorder are Gastro-Esophageal Reflux Disease and Irritable Bowel Syndrome. In approximately 40% of panic disorder patients, both are caused primarily though not entirely by the excessive and irregular secretion of bile so characteristic of hereditary panic disorder. In some patients the effects are mostly upward in the GI tract, resulting in excessive stomach acid secreted to neutralize the bile and leading to GERD – sometimes with associated asthma or sinusitis that is not recognized as related to panic disorder or GERD. (Gastric swelling and consequent pressure on the diaphragm or a nerve center has been proposed as one possible cause of abnormal breathing in panic disorder; such a suggestion is supported by the fact that subjective difficulty in breathing related to panic disorder can be alleviated in some individuals by pressing fingertips downward into the abdomen not far below the sternum.) Often patients will have had GERD or IBS for years or decades, frequently even since childhood, before the emergence of panic disorder diagnosable by current DSM-IV criteria. Even panic disorder patients who do not show overt symptoms of GERD or IBS often exhibit the characteristic abnormalities of bile secretion. The combination of Xanax and Carafate has been found to be remarkably effective in treating these conditions in panic disorder patients; indeed therapeutic response of GERD or IBS to this combination can almost be regarded as a diagnostic criterion. (It is not known why Carafate is often so much more effective in the panic disorder population than newer drugs like acid blockers and proton pump inhibitors, but the phenomenon is distinct.)
  
  
• Celiac Disease. Awareness of this previously neglected public-health problem is so new that we have no way of quantifying its comorbidity with panic disorder. However, it is obvious within the panic disorder community that CD or other forms of gluten intolerance are involved in a significant number of cases. Too many people improve on a gluten-free casein-free diet for the incidence of celiac disease in the panic disorder community to be as low as it is in the general population. (And too many patients come up positive for celiac disease when they finally persuade their doctors to order the immunological tests and the confirming endoscopy.) There is also the circumstance that on endoscopy of panic disorder patients, it is common to find opioid proteins "of unknown origin." They may be unknown to psychiatrists and perhaps to some gastroenterologists. However, celiac researchers are familiar with this finding and explain that the opioid proteins derive from gliadin and casein that have not been completely digested.
  
  
• Migraine. Estimates of the incidence of panic disorder in the population of migraineurs range from about 3X to about 10X the incidence in the general population – obviously a major association, whatever the exact figure, and an association some researchers claim is twice as strong as the association of migraines with depression or other anxiety disorders. One theory of migraine gaining favor among prominent researchers is that migraineurs are either situationally or congenitally impaired in the function of monoamine oxidase, an enzyme the body uses to "recycle" neurotransmitters and other peptides. Some of the peptides broken down by monoamine oxidase are powerful constrictors of blood vessels, and may be synthesized in the body or may come from dietary sources like cheeses, aged fish, and other dietary triggers of migraine. Monoamine oxidase inhibition would certainly qualify as a potential source of both biological and psychological hyperarousal.
  
  
• Throat, Bronchial and Sinus Problems. It is relatively common for panic disorder patients to have chronic problems of the throat, bronchia, or sinuses that are due to unrecognized "micro-spurts" of stomach acid up through the esophagus during sleep. Sometimes there are no symptoms of GERD other than the fact that such micro-spurts (or even merely the fumes from gastric hydrochloric acid) invade the throat, bronchia, or sinuses at night. A panic disorder patient with a robust, intact stomach lining does not necessarily feel gastric symptoms of GERD in such a case.
  
  
• Chronic Fatigue, Fibromyalgia. A growing number of researchers note the longitudinal association of chronic fatigue and fibromyalgia with panic disorder. It is not uncommon for someone who has experienced one of these conditions to develop panic disorder later in life. Typically such a person will continue to experience some of the symptoms of the earlier condition, but the expression of the ailment will become more clearly like panic disorder. It is very common for panic disorder patients to have some fibromyalgia-like pains. This association might lend some support to the controversial serotonin connection to panic disorder, although a serotonin role in chronic fatigue and fibromyalgia is not universally acknowledged.
  
  
• Depression. Depression is sometimes associated with panic disorder. Indeed primary care physicians appear most likely (probably due to intensive lobbying by pharmaceutical salespeople) to first try treating panic disorder with low-dose or even full-dose antidepressant medication. Some patients respond positively while others respond negatively – sometimes in a dramatically negative fashion. It is important to note that depression is often secondary to panic disorder, however. Panic disorder is a highly debilitating condition that often seems at first glance to have arrived quite suddenly and, if untreated or treated unsuccessfully or worsened by misguided treatment, can lead to acute disability and loss of employment. In many cases, one finds that what appears to be primary depression resolves of its own accord when the underlying panic disorder is successfully treated – even when the treatment did not involve the use of antidepressant medications or therapy.
  
  It is important to distinguish between primary and secondary depression and to consider treatment carefully, since the most popular classes of antidepressant medications reduce availability of acetylcholine and nitric oxide, both of which tend already to be at problematically low levels in panic disorder sufferers; indeed insufficiencient synthesis of acetylcholine and nitric oxide is implicated in recent research as an independent causative or exacerbatory factor in panic disorder, which may help explain why antidepressant treatment so often leads to dysautonomic crisis in panic disorder patients.
  
  
• Labile Hypertension, Benign Tachycardia. Hypertension that fluctuates widely, either during the day or from one day to another, should be a cue to screen for panic disorder. Tachycardia can also be an expression of panic disorder. Indeed a variety of arrhythmias, some benign and some not so benign, show an association with panic disorder. They are often difficult to observe, since they may occur only in conjunction with panic attacks or dysautonomic episodes. While it would make sense that beta blockers, alpha blockers, or alpha-2 agonists would be treatments of choice for panic disorder patients, real-world results with these medications are quite mixed. Panic disorder patients are often highly reactive to medications and experience more than their share of side effects and paradoxical responses. The first goal should be to stabilize the panic disorder and then to find an anti-hypertensive treatment that does not upset the equilibrium established in treating the panic disorder.
  
  
• Mitral Valve Prolapse Syndrome. Also diagnosed as "atypical mitral valve prolapse," or the symptomatology of MVP in the absence of organic evidence, this is sometimes a comorbid diagnosis or misdiagnosis for patients with panic disorder. It may be that MVPS is simply another member of the panic disorder, chronic fatigue, and fibromyalgia family, a member in which the most susceptible system is the cardiovascular system.
  
  
• Disturbances in Circadian Rhythm. Disturbances in circadian rhythm are prominent in anxiety disorders. Although the obvious interpretation is that anxiety makes it difficult for people to maintain regular sleep habits, there is also some evidence to suggest that careless sleep hygiene may just as well be a causative or exacerbative factor in some cases of panic disorder. Nocturnal panic attacks ("sleep terrors") are frequently present in certain variants of panic disorder, and tend to occur at the point of shifting from one sleep phase to another.
  
  
• Hypothyroidism, Hysterectomy. The author has not searched for and is not aware of relevant research [with later exception for the recent "chromosome 13 syndrome" research] to support more than a statistically predictable association of these conditions with panic disorder, but it is striking to see, among the panic disorder patients who show up in the Internet panic disorder support communities, how frequently hypothyroidism, often in one of its autoimmune forms, is associated with panic disorder. It is certainly well established that thyroid deficiency can cause symptoms of anxiety and panic. However, one also sees that some of the patients whose panic disorder might be attributable to a diagnosed thyroid condition have the "non-panic panic disorder" described earlier in this document. Another striking observation is the incidence of hysterectomy among female panic disorder patients. It is possible that primary care physicians are not adequately trained in the recognition and diagnosis of the various types of hypothyroidism. Certainly there is an abundance of panic disorder patients with hypothyroid symptoms who have only been tested with a TSH screen yet who have undergone emergency hysterectomies due to uncontrollable uterine bleeding – and whose physicians have never mentioned the connection between hypothyroid conditions and the type of uncontrolled bleeding that led to the emergency operations.
  
  
• Food Sensitivities. Food sensitivities are relatively common in panic disorder, as they are in MVPS as well. Some are attributed to IgG-mediated immune reactions; some are attributed to the alleged presence in grain proteins of constituent peptides that very closely resemble endogenous neuropeptides; some are attributed to irritation of the gastrointestinal tract and consequent depletion of neurotransmitters needed by the parasympathetic nervous system; some are unexplained. Some sensitivities may be as simple as those due to unstable blood sugar and highly glycemic foods. There is not nearly enough solid research in this area – nobody seems to know for certain (or at least to have proven) the mechanism of food sensitivities other than gluten or casein in panic disorder, but there is little question in the panic disorder community that even in blind situations, specific foods trigger dysautonomia and sometimes panic attacks in a subset of panic disorder patients.
  
  
• Allergic Reactions. Although panic disorder does not as a rule predispose a patient to allergies, immune reactions of any type, including IgE, IgG, IgA, and IgM, tend to increase the general level of autonomic arousal and thus have an association with an increase in panic susceptibility for the patient with panic disorder. The medical community is divided with regard to the classes of immune system mobilization that qualify for the term "allergic," but the fact remains that any significant mobilization of any part of the immune system, whether in response to true pathogens or in the form of a "false alarm," does increase autonomic arousal and the likelihood of panic attacks. A further complication is that the anticholinergic effect of antihistamines (frequently taken to combat IgE reactions) also increases the likelihood of panic attacks.
  
  
• Hyperstimulability. In some panic disorder patients, sub-paroxysmal dysautonomic episodes or even panic attacks can be triggered by flashing or flickering lights, or even by exceptionally bright lights. Rhythmic noise or high noise levels can have similar effects.
  
  
• Inner Ear Disturbances. Disturbances of the inner ear are common in panic disorder. Some clinician/researchers claim that such disturbances represent a causative factor in a substantial number of cases, citing the inner ear's close connection to the amygdala to support their claim. This would put treatment with SSRI medications into a new light, since these medications are themselves particularly rough on the inner ear. However, it remains to be seen whether inner ear disturbances prove in the end to be a widespread causative factor in panic disorder, a factor in a small minority of cases, or not much more than an interesting association.
  
  
• Volcanic Anger. In some panic disorder patients there is also a history of volcanic outbursts of anger. This can be a symptom of anxiety disorders generally, but it does seem to have more than a statistically random association with panic disorder. Patients should be asked about such anger in primary family members during an initial history, as it can give a clue as to the likelihood of an inherited predisposition to panic disorder. It is extremely common to encounter personal or familial denial of panic disorder, and volcanic anger can be a clue to such hidden cases.
  
  
• Alcohol Abuse. The abuse of alcohol is very common in panic disorder. It may take the form of binge drinking, in which case the panic disorder is probably not controlled or successfully hidden, and in fact the alcohol will be a factor in making the attacks more frequent. However, it is more common in panic disorder to see measured drinking, in which the patient drinks regular amounts at regular times of day, in effect self-medicating the panic disorder. In this case, the panic disorder may be successful hidden from other family members or the spouse. Sometimes the individual is not even aware of having the condition, having forgotten the initial panic attack or staved off an impending initial attack with alcohol. It has been postulated that one reason women are diagnosed with panic disorder more frequently than men is that men more often drink to cover the manifestations.
  
  
• Sexual Dysfunction. Sexual dysfunction is comparatively common in the panic disorder population. Panic disorder patients with parasympathetic insufficiency are predisposed to sexual dysfunction in the first place, and there is a good chance of ending up on one or more medications that will exacerbate the condition. Clinicians observe that panic disorder is hard on marriages even without the complications of sexual dysfunction. Every effort should be made to find a treatment plan that protects the patient's sex life.
  
  
• Schizophrenia. According to a recent meta-study, individuals who have a primary relative with schizophrenia are at equal risks of developing either schizophrenia or panic disorder. (The reverse does not hold true.) In addition, according to this meta-study, panic attacks and panic disorder often go unrecognized and untreated in patients with schizophrenia or schizoaffective disorders, exacerbating or sometimes mimicking the manifestations of the primary illness. It is suggested by some studies that in the minority of panic disorder patients with predisposition to schizophrenia, the emergence of panic disorder may presage or perhaps precipitate the development of schizophrenia, especially if undertreated or treated with medications carrying a risk for manic or hypomanic reactions.
  
  
• Epilepsy. It is becoming increasingly clear that epilepsy and panic disorder coexist more frequently than chance alone would predict. There have been numerous neurology papers and discussions about this in recent years, and there is a growing trend to try antiepileptic medications in the treatment of panic disorder even when epilepsy itself is not present. There is potential for misdiagnosis in either direction (seizure / panic) of a particular episode, but there are clearly patients who experience both seizures and panic attacks distinct from each other. The question is primarily the frequency with which such patients are to be found. The model of a panic attack as ectopic firing of neurons peripherally rather than within the brain, a model posited to apply to some but not all cases of panic disorder, is congruent with the accepted model of seizures. It is interesting that some of the same triggers and treatments, both medical and dietary treatments, apply to both epilepsy and some cases of panic disorder even in cases where they do not coexist.
  
  
• Autism-Spectrum Disorders. One of the most striking associations on clear display in panic disorder support groups is the familial association between panic disorder and autism-spectrum disorders. Sometimes both diagnoses reside in the same person, but far more often it is a parent or sibling who has panic disorder and a direct relative who has the autism-spectrum disorder.

As one follows any group of panic disorder patients over a period of years, half or more do apparently keep the diagnosis of essential panic disorder. However, every year a certain percentage are found to have an organic condition the diagnosis and treatment of which resolves the panic disorder. Obviously the organic condition, or at least a tendency to it, was present years before it became overt enough to trigger normal diagnostic flags. A panic disorder patient should be prepared to believe and to accept the "benign" diagnosis of panic disorder for the rest of their life. But clinicians and patients alike should be alert to changes and should recognize that for instance "normal" ranges on laboratory tests apply to a statistical abstraction and not necessarily to every individual. Often individuals experience well-being only within a narrower range of values than the stated range (which varies from lab to lab in any case), and on some types of tests, even the optimal center point may be different for different individuals.

Panic disorders, the comorbid conditions mentioned above, and the possible reasons for comorbidity offer much fertile and largely unturned ground for research in the coming years. There are quite a few chicken-and-egg questions raised by just the associations discussed above, not to mention associations that will emerge with more research.

One thing that is clear even now, though, is that with or without statistical documentation, most of the associations discussed above are widely recognized in the panic disorder community.

Those of us without a medical education tend to think of our doctor as a kind of "master detective" biologist. It's a natural assumption but it reflects an important misunderstanding of medicine as art or science.

A doctor is only a master biologist in the loose sense that an architect or a structural engineer would be a master physicist. After a few notoriously rigorous courses in certain traditional elements of human biology, a medical education branches rather quickly into the study of established medical protocols for diagnosis and treatment of well-recognized conditions. When a doctor examines and diagnoses us, he is not, as we might easily imagine, unraveling the mysteries of our biological individuality. Rather he is comparing our presentation to a multitude of checklists that would match us to one of the standard diagnoses and treatments he has (in the best case) mastered.

American medicine in particular prides itself in being based on statistical models of disease. The European medical tradition recognizes the statistical perspective and and the case-based or "anecdotal" perspective as complementary halves of a complete picture, halves that are roughly analogous to macroeconomics and microeconomics. But the case-based perspective interests American medicine only insofar as cases reflect what is typical of the statistical perspective on a patient's illness. Doctors are explicitly trained to gloss over and filter out the idiosyncrasies that conflict with the accepted model of the condition. Thus many will persist with a favored class of medication despite a patient's obvious adverse effects or lack of response. It is revealing that the doctor's impatience will be with the patient, not with the medication. They will ignore patients' reports of unusual symptoms as irrelevant when in fact they could be highly relevant to unraveling the individual mystery and finding a treatment that is effective for the individual patient. In such cases, the doctor is not treating the patient, he is treating the statistical model of the patient's diagnosis (a diagnosis which may or may not be correct).

Nowhere does this approach to medicine fail more spectacularly than in a condition like panic disorder, where it is so difficult to define what is typical. But it fails in other contexts as well.

The author of this document himself was evaluated by 14 doctors over a span of more than 10 years before aldosteronism was finally noticed, proven, and identified as the root cause of all the symptoms that had received such a variety of diagnoses and misdiagnoses (including panic disorder) over that time. Mumblings about "anxious patient" and "panic attacks" were always somewhere in the background. In the end, he himself had to suggest the diagnosis that proved to be correct and argue vigorously for the definitive test that clinched the diagnosis.

Unfortunately, the author's experience is not at all unusual. Such outcomes are commonplace in panic disorder support groups. They may even be more the rule than the exception. The overlooked diagnoses vary, but the experience is the same. Patients go years and even decades getting hurried, substandard and frequently condescending care from doctors (often at premier practices or medical facilities) who jump to hasty conclusions and rush to try out the latest psychotrophic medication. Often the final and correct diagnosis is a condition thought to be rare, such as celiac disease, aldosteronism, or thyroid dysfunction, but which appears to be strikingly common in the "panic disorder" population.

Sometimes these failures of diagnosis occur because a doctor who might recognize a single unmistakable result on a routine test or some other unmistakable clinical indicator is not good at recognizing a constellation of borderline results or suggestive indicators that must be considered in combination before they add up, as in the author's case, to a clear picture.

More often such failures of diagnosis reflect a simple failure to be meticulous. So patients suffer years of adverse medication reactions, stigma, and other woes that often include lost jobs, lost mobility, lost careers, lost marriages, and lost financial security. Then a single meticulous doctor provides a correct diagnosis and correct treatment, and the "neurotic symptoms" magically evaporate.

One truth is abundantly clear in all of this: the extent to which our presentation is atypical is almost precisely the extent to which our doctors will fail to correctly diagnose and treat us. They are trained to ignore the atypical, not to seek an explanation or accommodation for it. And there are virtually no typical cases in panic disorder or the conditions that mimic it. There are only typical treatments.

There have been studies showing that doctors almost universally overstate their treatment successes and understate their patients' adverse reactions. Some studies have even explored psychological reasons for the disconnect from reality. But it is also easy to explain by natural selection. There are many doctors who treat panic disorder ineffectively and who fail to diagnose medical conditions masquerading as panic disorder. There are only a few who routinely do well at it. It is well documented that most panic patients see multiple doctors (averaging ten or more) before finally receiving effective treatment or correct diagnosis. Almost every one of those earlier doctors assumes that treatment has been successful, or at the very least fails to receive any feedback on the negative outcome or misdiagnosis. Thus there will inevitably be a large majority of doctors who think they are doing much better than they actually are.

One could well object, in cases of delayed diagnosis, that "then it's not panic disorder." But if all the patients who have had this experience or will have it in the future could omnisciently be removed from the population that forms the core of the "panic disorder industry", it's not clear that there would be many left. Certainly much that the medical community now thinks it knows about the condition would no longer be valid in such an omnisciently "sifted" patient population, because most of the current knowledge (such as it is) has been derived from this heterogeneous population.

Unfortunately, all of this talk about undiagnosed and misdiagnosed conditions feeds into the worst and most common fear in the panic disorder community, namely that some terrible condition has gone unnoticed, one that will kill us before it is discovered. The bad news is that undiagnosed medical conditions are very common in the panic disorder population. But the good news it that they are almost never fatal conditions. The fatal conditions do get diagnosed quickly. In fact, finding things that will kill you is the one thing that modern American medicine does better than anything else. By the time a "panic disorder" patient has undergone EKGs, MRIs, and all of the other routine technological marvels requiring little initiative or detective work from a doctor, the patient can be confident there is no undetected brain tumor, no undetected adrenal tumor, no potentially fatal heart arrhythmia, and no other killer lurking unseen in the background. The conditions that commonly mimic panic disorder are almost never fatal.

But it is inarguably difficult to get adequate testing. Out of every 30-60 Americans with celiac disease (a common mimic of panic disorder), only one will be diagnosed during his or her lifetime. Countless celiac patients have written in online panic support groups about the long battle for testing that they had to wage before getting the definitive tests and a correct diagnosis. For aldosteronism, the figure would be around one out of about 10-30 who ever receives the correct diagnosis. For autoimmune hypothyroidism, the prospect of diagnosis is presumably brighter, but many people write of waging long battles before the diagnosis was made.

Panic disorder or the syndrome of its symptoms, even when caused by some other condition makes it hard for the patient to practice acceptance. Yet it is vital to do so. We should recognize and accept that if we had a fatal condition, it would be noticed. We should recognize and accept that there may very well be some non-fatal condition causing our symptoms that has not yet been noticed – and that is not a catastrophe. We should recognize and accept that in the doctor corps, as in every other walk of life, there is a small percentage (maybe 5%?) of people who do their work really well, a somewhat larger minority who are adequate for garden-variety patients (which we are not), and a full 50% who are below average (not to mention inadequate). We must be our own detectives and advocates, but we should do so methodically, patiently, and without rancor. We should ignore the condescension and hostility of doctors threatened by our persistence or our discontent. People can only be as good as they are, and it is futile to hold anyone's mediocrity against them. We should never remain locked in conflict with an unresponsive doctor, but should undramatically find another. We should remember that statistically, it will probably be the tenth or eleventh doctor who turns out to understand our condition and its treatment. Every new doctor is a step toward the right one. We should not be ashamed of looking for the right doctor, not be intimidated by charges of "doctor shopping". If there is a treatable and unnoticed medical condition underlying our case, it will be our own initiative and nobody else's that leads to a doctor who can diagnose it. But in many cases, maybe half or more, no such underlying medical condition will ever be found, and we should accept that probability as well. The fact that panic is so hereditary (and genetically linked to a variety of other conditions) makes it clear that biology is a strong factor, but not every biological problem has found its way into the protocols of medicine. We should accept that as well. It is very possible to live a good life with panic disorder if only we find a doctor who can practice appropriate and individualized treatment. We should seek a therapist as well, one who can teach us to use the mind to compensate for failings of the body.

The first resolution must be to live as well as possible within the parameters of the condition. Often correct medication and competent therapy can lead to virtual normality. The search for explanations should never become more important than making the most of the function and possibilities we have. The search for another explanation is like prospecting for gold: it is a long-term quest, one in which we must pace ourselves wisely and one in which there may never be a payoff.

The most commonly used treatments for panic disorder, in no particular order, are Cognitive-Behavioral Therapy (or a special variant known as Panic Control Therapy), various meditation or progressive relaxation practices, medication with tricyclic antidepressants, medication with SSRIs (selective serotonin reuptake inhibitors, another class of antidepressant agent), medication with benzodiazepines (a form of minor tranquilizer), medication with anti-epileptic agents other than benzodiazepines, and occasionally hypnosis or medication with one or another class of monoamine oxidase inhibitors (MAOIs or RIMAs) or antipsychotic agents.

Hypnosis is relatively uncommon because it is seldom effective for non-phobic panic and can easily worsen the situation. Attempts to unearth and treat the "underlying trauma" are usually misguided because if the initial diagnosis of panic disorder has been rigorous, it appears that there is seldom an underlying trauma to unearth.

Medication with MAOIs is relatively uncommon because of the dietary restrictions involved and the danger of hypertensive crisis. Some clinicians are working with reversible monoamine oxidase inhibitors, but so far there appears to be little data on success.

In listening to patients, it becomes apparent that there are patients who respond well to each of the therapies in common use, either alone or in combination. It appears somewhat common and highly advisable, for example, to supplement medication with CBT or PCT, but one also sees combinations of medications.

Reactions to medications appear to be highly variable in panic disorder – less consistent, for instance, than reactions to anti-hypertensive medications in the treatment of hypertension. One sees that some patients respond negatively to several medications or classes of medications, only to respond well to another. Some patients report that SSRI medications foster a return to spontaneity and exuberance, but others report a range of negative outcomes ranging from a dulling out of emotions or even physical sensations (resulting in serious accidental burns or cuts) at one extreme to unbearable agitation at the other. Similarly, some patients report a normalizing effect from benzodiazepines, but others experience negative outcomes ranging from stupefaction or emotional dulling at one extreme to awkward or even dangerous disinhibition at the other. It seems quite clear in talking to patients that reactions are highly individual, and it seems probable that this is due to differences in respective patients' underlying biochemical disturbances.

(A number of studies have pointed out that generally speaking, antidepressants seem to work well when the primary manifestations are of a psychological character, while benzodiazepines seem to work better when the primary manifestations are of a somatic character. Here again we see evidence that there are separate etiologies underlying such a systematic difference in efficacy – that indeed distinct conditions are probably involved: a sort of "GAD plus panic attacks" and a "dysautonomia with paroxysms and sometimes secondary worry.")

It appears that few psychiatrists are aware of the distinct varieties of panic disorder outlined in the beginning of this essay, although the research documenting them is readily available in research journals and correct identification of the variety of panic disorder is critical to the correct choice of treatment. Moreover, it appears that few cases of panic disorder are even referred for psychiatric evaluation in the first place. Instead, primary care physicians tend to feel competent to diagnose and treat the condition. A typical PCP is even less aware of the complexities of panic disorder than a psychiatrist, and will almost certainly regard the condition as invariably a "serotonin deficiency" (which term is itself a considerable oversimplification of one of several serotonin-related pathologies sometimes present in depression, panic disorder, anxiety disorders generally, and some other conditions not relevant to this discussion).

So while it is quite apparent that reactions to panic disorder medications and the correct choice of medication vary by individual and by subtype of panic disorder, it also becomes apparent in listening to patients that many if not most physicians prescribe not according to what should or does work for individual patients, but according to what "works" for the individual doctor. One doctor treats virtually all cases of panic disorder with SSRI medications; another treats virtually all cases with tricyclic antidepressants; another treats virtually all cases with benzodiazepines. Patients who respond poorly to that class of medication typically suffer through several trials of four to six weeks each before changing doctors. It is not uncommon for a panic disorder patient to have gone through a number of physicians (with considerable personal suffering and sometimes loss of employment) before responding well to treatment. Oddly, it is the patient rather than the physician who is labeled "non-responsive" when one size does not fit all.

(Patients also tend to be unaware of the large proportion of doctors who have conflicts of interest in the form of subtle and sometimes not-so-subtle financial incentives to prescribe particular medications. In this respect, the ethical canons in medicine appear, at least based on actual practice, to be remarkably lax in comparison to those of other licensed professions.)

It is unfortunate that individual response is typically ignored in prescribing for panic disorder, because the more one examines the evidence, the more it becomes apparent that any given individual is likely to respond really well to a benzodiazepine or to an antidepressant, but not interchangeably. Response to one of these classes of drugs is likely to be optimal, while response to the other will be marginal at best. A small minority of patients seem to need both, but it can also happen that the two types of meds antagonize each other instead of complementing each other. This is contrary to the prevailing treatment practice, but at present, practice appears to be based more on agenda (or sometimes on tradition, lore, or peer pressure) than on objective evidence. Indeed most non-specialists appear to be wholly unfamiliar with the actual evidence, which is often at odds with the claims of the salespeople (most of whose education is in sales,marketing, or communication) on whose counsel they rely.

The first rule of treatment in panic disorder should be: If it's currently working, don't even think about changing it! It is utterly appalling to see the numbers of patients who have been controlled and highly functional with one strategy of management over the course of years, only to move to a new state or change employer (and thus change insurance company) or have their doctor retire, and then have complete havoc wrought in their medical condition, their emotional state, their marriage, and even their employment and financial status, all because an inept and overly self-assured doctor sought to adhere to one or another of the continually emerging vogues in prescribing. One need only search MedLine to see the course of these successive vogues over the last two decades, and it is clear that what is "in" today will be "out" a few years from now. There are choices to make in treating any newly emergent case of panic disorder, and humane, flexible, patient-responsive experimentation may be necessary. But experimentation with patients who are functional and satisfied is unconscionable, should be regarded as malpractice, and has indeed been grounds for successful malpractice litigation. Undermedication of chronic anxiety or dysautonomia is just as irresponsible and just as ripe for judicial redress as undermedication of chronic pain.

A large proportion of patients undergoing CBT or PCT report that they have experienced significant benefits from such therapy, particularly in the area of coping with their symptoms, reducing the severity of their panic attacks, and improving their overall well-being. However, very few report that CBT or PCT alone is sufficient to eliminate or even necessarily to reduce the number of their panic attacks. Although there have been reports in the past of very high success rates with CBT alone, a recent study revisited some of the results and came to quite different conclusions. David Barlow, Ph.D., one of the leading figures in psychological treatment of anxiety, commented on this study with the laudable observation that "The one thing that we've learned ... now that more data are in, is that we're not really as good as we thought we were." This was refreshing candor and integrity in a field where professional pride has often taken priority over a dispassionate and monetarily disinterested look at the relative merits of various approaches to treatment.

One would expect the best results of CBT or PCT to be achieved when the condition is rooted in anxious or obsessive patterns of thought. However, it is not always easy to distinguish such cases. Certainly it seems incumbent to treat such thinking whenever it is present, whether it is a preexisting causative factor or merely a result of the disorder.

Further investigation is needed into the effects of CBT and PCT on the dysautonomia associated with panic disorder. Although a patient is likely to be very grateful (and at a sharply reduced risk for suicide) after learning to experience less severe panic attacks in the face of prolonged dysautonomia or dysautonomic paroxysm, little has been gained from a medical perspective if the patient nonetheless experiences prolonged and frequent dysautonomic episodes. There is a certain value in teaching patients to cope with the symptoms of medically uncontrolled essential dysautonomia (especially if such treatment may reduce actual symptoms), just as there is value in teaching patients to cope with the symptoms of other chronic medical conditions. Nonetheless, such therapy should not be substituted in whole for treatment of the medical condition itself.

Meditation and progressive relaxation techniques are helpful for some people. Meditation techniques need to be undertaken with the understanding that dissociative states like those it may promote are not always desirable in panic disorder patients – that indeed they can trigger panic attacks in themselves or simply be more difficult to terminate for individuals with anxiety disorders. (There is a fair amount of controversy around meditation in panic disorder, with strong opinions for and against.) And individuals undertaking either meditation or progressive relaxation need to understand that even when the practices are beneficial, a panic attack is most likely to occur during the transition from a state of high arousal to one of lower arousal. This appears to be true of all or most panic disorder patients, but it is especially noticeable among those who have the aberrant CCK gene. Obviously it is desirable to make the transition from high to low arousal, but most people would prefer to do so without triggering panic attacks in the process. Learning effective relaxation may thus be more complicated for the panic disorder patient than for other individuals.

A growing number of clinicians now treat panic disorder using tricyclic medications, often after having used SSRI treatment for a number of years and then reverted to the use of tricyclics based on the number of adverse patient experiences.

Tricyclics seem to work well for a minority of panic disorder patients. Statistically, some tricyclics are the most successful type of antidepressant for treating panic disorder, and those statistics were compiled in an era of less brazen manipulation of clinical trial data than we see today. However, due to the parasympathetic insufficiency so frequently present in the disorder, a certain subset of panic disorder patients seem particularly prone to excessive cardiac stimulation from this class of medications, which increase the availability of norepinephrine. Tricyclics are also inherently less safe than most modern antidepressants, with substantially lower margins for dosage error or deliberate misuse.

Tricyclic medications also reduce the availability of acetylcholine, and reduced availability of acetylcholine is believed to be implicated in the mechanism of panic disorder. Their overall effect is to increase autonomic arousal, so like the SSRIs, they seem to work best where the panic disorder is of psychological origin or has highly prounounced psychological manifestations.

Among clinicians new to treating panic disorder or new to using SSRI-type medications, these medications are often regarded as the treatment of choice or even a panacea. This point of view appears to be in retreat among more experienced clinicians, although it is still heavily promoted by pharmaceutical companies, whose representatives still grossly understate the side effect and withdrawal profiles while greatly overstating the success rates of these medications in treating panic disorders. However, the side effect profile for SSRI medications is much worse than initially published studies revealed, the effectiveness is highly unpredictable in treating panic disorder, and there is high risk for precipitating iatrogenic crisis when using this class of medications to treat panic disorder (especially during the beginning of treatment).

An alarming percentage of patients in the Internet support communities for panic disorder had never experienced or even heard of a panic attack before they were treated for depression with an SSRI, but experienced panic attacks leading to an ultimate diagnosis of panic disorder within their first week or even the first day of treatment and have not been able to rid themselves of the disorder after discontinuation of the SSRI. Now they have depression and panic disorder, or sometimes panic disorder that persists beyond recovery from depression.

Primary care physicians on the whole, and a sizable minority of psychiatrists, are unfamiliar with the true side effect profiles of various SSRI medications, are unaware of the tolerance that develops to these medications when they are used to treat panic disorder, and are unaware of the withdrawal syndromes, despite the fact that all of these have been well documented in such sources as the British Medical Journal and the Journal of Clinical Psychiatry, among others. (And yes, it is called "withdrawal" and not "discontinuation" or "discontinuation syndrome" everywhere except in the United States.)

A disturbing number of patients are not informed of the potential long-term risks of their medications, including the significant weight gain and sexual dysfunction, both of which can persist long after discontinuation of SSRI medications. When patients report some of the side effects of usage or discontinuation that are typical in panic disorder, such as visual lag and sensations of random electric shock, their physicians incorrectly state that these are not recognized as attributable to SSRI medications, even though they are well documented in the psychiatric literature.

Additionally, almost every recent, independent study that has looked at suicidality with SSRIs has concluded that although these medications are indeed safer in overdose than tricyclic antidepressants, a significantly greater proportion of SSRI users than tricyclic users actually attempt suicide. These people are not uninformed, either – they know that their SSRI is not lethal, so they tend to use acetaminophen, gunshot, or hanging instead.

Both anecdotal reports and systematic surveys of clinical practice reveal that a majority of physicians prescribing SSRIs do not know how to correctly taper the beginning and end of treatment with these medications, nor what reduced dosages to use in treating panic disorder, nor what protocol to follow in switching from one SSRI to another – usually a direct switch is advisable, but some physicians prescribe a washout period that is both uncomfortable (if not debilitating) and dangerous to the patient – nor how to switch between an SSRI and another class of antidepressant.

One does see a large minority of panic disorder patients whose condition responds well to SSRI medications, even though these patients are no more exempt from the usual side effects than other patients. (Some find that the addition of Serzone [now banned], or less often Wellbutrin, can restore sexual function lost to the primary SSRI and/or reverse the perversely treatment-resistant weight gain experienced by a majority of long-term SSRI users; however, neither of these medications appears to be useful alone in the treatment of panic disorder.) Not all of the patients who respond to SSRI treatment are the ones in whom the condition seems attributable to anxious or obsessive thought patterns, so it appears likely that there is some set of patients for whom these medications correctly target some underlying abnormality or imbalance. However, many physicians seem unaware that a significant number of panic disorder patients react very badly and sometimes very suddenly to the anticholinergic effects of SSRI medications, a reaction that should not be surprising given the parasympathetic insufficiency posited to be so often present or even causative in panic disorder.

Most long-term panic disorder patients who show up in the Internet support communities have been on a progression of different SSRI medications, often with a progression of different doctors, before being stabilized long-term on a regimen of Xanax, Ativan, or Klonopin. A small number end up on Valium, Tranxene or other benzodiazepines. [Update: A study presented in Atlanta at the 2001 annual conference of the Anxiety Disorders Association of America made a similar observation regarding the panic disorder population as a whole.]

A significant number of other panic disorder patients who show up in the Internet support communities began treatment on both an SSRI and a benzodiazepine, but were forced to discontinue the SSRI due to unacceptable side effects, and then found that they were well controlled on the benzodiazepine alone.

One of the disturbing characteristics of SSRI medications is that once a particular SSRI has ceased to be effective in controlling panic disorder, it usually cannot be used effectively again in the future, not even years later – the cessation of effect is persistent and perhaps permanent. This raises some troubling questions about the mechanisms behind the well-known "poop out" phenomenon distinctive to SSRIs and other reuptake-inhibiting medications. Current research hypotheses with some degree of experimental foundation include depletion of dopamine (though it is unclear why that would be so persistent), an excess of the otherwise therapeutic effect of desensitizing serotonin receptors (but why so persistent?), and physiological damage to or destruction of serotonin receptors (in which case it is doubtful but unproven whether they can be regenerated).

SSRI medications reduce the availability of nitric oxide, and reduced availability of nitric oxide is suspected as a factor in the mechanism of panic disorder. They also increase production of prolactin, which may be a factor in the high incidence of sexual dysfunction, and alter the function of the hypothalamus, which is one of the mechanisms being studied as a possible factor in the strangely treatment-resistant weight gain associated with SSRI usage.

(It is a shame that the dramatic differences in reactions to SSRI treatment are so often minimized, trivialized, or dismissed outright instead of being acknowledged and studied more closely. It is a shame because the differences probably provide, in their sheer starkness, some of the best clues currently available to the varieties of biochemical mechanisms underlying panic disorders – and to the various mechanisms underlying depression and obsessive-compulsive disorder too, for that matter. Groups of people with similar reactions to one class of drugs, particularly drugs as selective and powerful as SSRIs, probably represent de facto classifications of people with similar variants of panic disorder or depression. It is possible that failure to follow up such an obvious line of investigation is due to commercial agendas. But it is also possible that it is simply easier, with current research technology, to plan and study interactions between drugs and neurons in the laboratory than it is to look inside a living human being and observe the individual's true neurochemical processes or status at a given moment – that our ability to create drugs that act at the neuronal level has far outpaced our ability to observe in thorough and objective ways their actual effects on living creatures' neurochemistry. And so we rely on a priori speculation, we rely on subjective reporting of sensations and feelings far removed from the biochemical processes that cause them, and we conjecture as to their meaning. Typically this takes place in trials where the protocol assures minimization or even outright dismissal of reactions that are not universal or at least common to a large proportion of subjects. So when something goes wrong for a minority of patients, we conjecture around variations on the presumed and intended workings of the drug at the biochemical level; but we do not know, and we have few tools to uncover, the subtleties that explain the full range of unintended effects.)

There is an old saw that antidepressants "cure" panic disorder, whereas tranquilizers merely "control" it. Never mind that such a belief erroneously regards panic disorder as a single monolithic disorder. Even overlooking that point, when one looks beyond the short-term perspective of typical clinical trials, it becomes apparent that antidepressants do not "cure" panic disorder, other than in rare cases – the condition is chronic and often mildly progressive. Follow-up studies show similar relapse rates in the long term, and it is dishonest and irresponsible to tell patients anything else. With CBT or PCT, one can sometimes reduce the incidence of outright attacks by reducing sensitivity to triggers, but the underlying dysautonomia persists – and it is the dysautonomia, not its paroxysms, that represents the most insidious threat to health and longevity. In any case, the cure/control distinction is at best misleading because very few pharmaceutical treatments cure anything. Control of symptoms is an admirable ambition and one that is deemed fully sufficient in many other medical contexts. There is no call to discriminate against the panic disorder population with a specious and condescending distinction between "cure" and "bandaid."

The fixation on SSRIs as the one-size-fits-all treatment of choice for panic disorders is largely attributable to a marketing campaign targeting primary care physicians and their patients that was conducted for SmithKline Beecham (as the makers of Paxil were known at the time) by the Ruder Finn agency. As Ruder Finn trumpeted until recently on a World-Wide Web site intended for advertising industry insiders, the "Paxil for Panic Disorder" program was one of the most spectacular successes of pharmaceutical blitzmarketing in the history of the industry. To quote from the Web site:

The Paxil for Panic Disorder program leveraged a new indication of an existing antidepressant for a relatively obscure anxiety disorder to generate "unprecedented" product sales growth. Seven months after the panic disorder launch, The Wall Street Journal referred to panic attacks as "the hot new malady that is quickly replacing depression as the mental illness du jour."...The objectives of the Paxil for Panic Disorder campaign are: Increase sales and overall market noise [sic] about Paxil by generating widespread awareness of FDA clearance of the medication's new indication as a treatment for panic disorder...Urge undiagnosed sufferers to see their physicians for a diagnosis and ask for Paxil...Take advantage of seasonal and other opportunities...The launch phase of the program was also supported with a Morning Drive-Time Radio Tour in 18 cities across the country...A satellite media tour was considered [but dropped]...Tie in to the stressful nature of the holiday season...

Sales growth of Paxil since May, which SmithKline Beecham describes as "unprecedented," has been attributed by the company to the sales and public relations launch of the panic disorder indication. SmithKline Beecham saw its share of the total SSRI market share grow by 8.7% and its share of new prescriptions grow 7.7% between April and November.

Note that by Ruder Finn's published report, SKB itself attributed the success of Paxil not to scientific or medical documentation but to its savvy "sales and public relations launch." The clinical trials submitted as the basis of the FDA approval and ensuing media blitz were three, and in them the drug was successful in eliminating or reducing panic attacks in 25% (misstated as 76%*), in 51%, and in 33% of trial subjects on Paxil, versus 44%, 32%, and 14% of subjects on placebo. To specialists in the treatment of panic disorder, these numbers are anything but impressive – indeed if the drug had been promoted on numbers like these rather than the sweeping generalizations primary care physicians hear from pharmaceutical salespeople, it is unlikely that much enthusiasm would ever have developed in the first place. Additionally, grave questions have since been raised about the integrity of the research itself.

* The company did not cite a figure for all subjects in the first study, but instead counted only the one-third of subjects who responded best – and who were on a dosage intolerable to many patients. Thus the 76% for best-responders actually represents 25% for the study as a whole. (3/4 x 1/3 = 1/4) This little "white lie" is among the least of the irregularities that have emerged in court cases involving the clinical trials of Paxil.

The Ruder Finn campaign aimed to broadcast the "cleaner side effect profile of Paxil," although the sad truth is that clinicians are largely in agreement today, about three years later, that of all the commonly prescribed SSRIs, Paxil seems to have the "dirtiest" long-term side effect profile and the most problematic withdrawal. (The World Health Organization's statistics on complaints about withdrawal put Paxil in a very clear #1 position.) Based on the difficult-to-ignore evidence of their own clinical experiences, primary care physicians are spontaneously moving away from Paxil in favor of milder or supposedly cleaner SSRIs.

The plaintiffs in a California class-action lawsuit against the makers of Paxil (seeking changes to labeling, release of embargoed data, and only modest monetary damages) intend to introduce documentary and testimonial evidence: that Paxil performed significantly, indeed markedly, worse than placebo in its early trials, before statistical spin doctors were brought in to fix the problem; that when dozens of trials did not yield positive results, statisticians manipulated the numbers to produce "hypothetical" trials in which one single patient who endured or thrived on Paxil for 365 days would completely offset 365 patients who could not endure the medication for more than one day; that the makers changed numerous reports (adding up to tens of percents) complaining explicitly of withdrawal, as distinct from relapse, to instances of "relapse" instead; that the company suppressed evidence of a 0.77% incidence of suicidality (considered quite high for such a serious side effect, and later demonstrated to be even higher); that a certain identified FDA official illegally coached the makers of Paxil in clandestine conversations on how to avoid scrutiny on the suicidality issue; that the company suppressed information on a 20-25% dropout rate in early trials; that the company deliberately claimed on the record in an FDA hearing that it had studied withdrawal issues and found none, when in fact it had not studied withdrawal issues but had inadvertently become aware of frequent and severe withdrawal experiences; and that the company has systematically promoted the impression that the only people who become dependent on Paxil are multi-drug abusers, even though the company has abundant evidence (which it suppresses) that dependence and a pronounced neurological withdrawal syndrome are common. As internal company documents have come to light in other civil and criminal trials regarding Paxil in the United States and other countries, it has become clear that the drug's makers possessed far more documentation of the drug's potential for adverse effects than they ever disclosed to the FDA, to the public, or to the scientific community.

There is obviously a place for the discriminating use of SSRI medications in the treatment of panic disorder. They are irreplaceable, indeed life-saving for many people. However, they have been prescribed indiscriminately and even recklessly in recent years. The appropriateness of their use has been greatly exaggerated and the long-term risks and sometimes severe dependence and withdrawal effects (which bear no resemblance at all to the condition being treated) have been swept under the carpet for as long as humanly possible. The question is not whether there is a place for SSRIs in the treatment of panic disorder. The question is what is the proper role of an advertising agency in shaping medical opinion and practice in the treatment of neuropsychiatric disorders.

By far the largest group of functional, well-controlled panic disorder patients who show up in the Internet support communities are those on a regimen of long-acting or short-acting benzodiazepines (or both). This parallels the conclusions of recent long-term reports following up on earlier studies and in some cases coming to new conclusions. The most common regimen among such patients is a long-acting benzodiazepine taken on a regular schedule, with a short-acting benzodiazepine to be taken for "breakthrough" symptoms. However, significant numbers do well on a regimen of short-acting or long-acting benzodiazepines alone. A substantial proportion of these patients take only enough medication to manage their autonomic arousal, thus breaking the autonomic link in the arousal-anxiety-arousal-panic cycle, and normally rely on techniques learned in Cognitive-Behavioral Therapy to manage the cognitive-emotional aspects of the condition.

Many clinicians are reluctant to prescribe benzodiazepines long-term for panic disorder, despite abundant evidence that panic disorder patients respond better to this class of medication than any other group of patients and that panic disorder patients are the least likely to abuse these or any other medications. The typical pattern is that panic disorder patients require some adjustment of dosage in the beginning of their treatment – either up or down – and then remain at or below that dosage for years. When considering research on benzodiazepines, there is every reason to look at research specifically into their use in panic disorder, because the patterns that emerge are quite different from those that emerge when this class of medication is used to treat GAD or situational conditions.

It is common for panic disorder patients spontaneously to reduce their medication after six months to a year, and to remain satisfactorily controlled at the lower maintenance dosage.

It is common for panic disorder patients to try to eliminate their medication after a considerable period free of symptoms, only to find that they must go back on their original level of medication. There may be a medication-free period of success lasting months or occasionally years, but the condition eventually returns – sometimes with an intensity requiring more medication than before. Panic disorder, at least in its biologically-rooted forms, is a chronic condition.

Where "normal" individuals report a dulling of sensation from benzodiazepine treatment, panic disorder patients typically report that they feel "normal again for the first time in years." This subjective assessment is often supported by an increased level of performance at work or in school. Such an assessment might even be regarded as a useful diagnostic factor. The baseline levels of autonomic arousal and of parasympathetic nervous system activity tend to be different in panic disorder patients from those found in normal individuals, and that helps explain why benzodiazepine treatment can be not merely therapeutic but normalizing in the panic disorder patient, with a lower incidence of side effects than those experienced by the general population and an incidence of abuse so low as to be almost nil.

Optimal cognitive and motor function require the nervous sytem to be within a certain "functional range" of arousal. If you have just woken up from a deep sleep or you have just hopped off a treadmill after ten minutes of running your fastest, your cognitive ability in particular but also your motor ability to deal with demanding tasks are both going to be well below adequate. Your nervous system is outside its functional range of arousal. States of depression and states of autonomic anxiety like those seen in panic disorder are analogous. Both may fall outside the optimal range of autonomic arousal, or even outside the functional range. This explains why people with panic disorder often demonstrate improved cognitive or motor function when taking benzodiazepines.

Learning is a particularly demanding task, and the brain is biochemically inhibited from learning well when there are high levels of arousal-inducing hormones or neurotransmitters. Thus patients with panic disorder or a condition that mimics it may be unable to learn in an academic setting and unable to benefit from Cognitive-Behavioral Therapy (which teaches the brain new patterns of thought and new reflexive behaviors) until an appropriate dose of a benzodiazepine is taken. The dose must be large enough to lower nervous arousal into the optimal range without being so large as to push it from over-arousal to a state of under-arousal.

Much of the early research on the use of benzodiazepines in anxiety and panic was conducted with dosages that today appear stunningly high. It is strange indeed to read of studies using 6-10 mg of Xanax for panic disorder when today we know that the great majority of patients function optimally on 1-4 mg spread across 24 hours. Some patients require as much as 6 mg per day, but that is so rare that even the most benzodiazepine-friendly physician would almost certainly question the diagnosis or consider a different type of medication. At present Xanax is the best-known benzodiazepine used against panic disorder (and the one that is FDA-approved for the condition), but perusal of early research into the use of Klonopin and other benzodiazepines against panic disorder shows the same thing: the dosages tested tended to be far higher – 2x, 3x, or even 4x – than the dosages commonly used and known to be effective today.

It is of course important to taper the cessation of benzodiazepines carefully. Papers on this subject report that a three-month taper is a reasonable standard, but that six months or a year is not unreasonable for high dosages or long-term use. In the case of panic disorder, there should not be undue pressure to get off of any class of medication, and during cessation it is often difficult to distinguish withdrawal symptoms from symptoms of the underlying panic disorder (except in the case of discontinuing SSRIs, where the most dramatic withdrawal symptoms are clearly neurological and bear no resemblance to panic disorder or depression). A number of papers report that panic disorder patients are frequently withdrawn successfully from their medication only to have to go back on it at the same or a higher level half a year or a year or two later. These papers suggest that it is probably healthier to maintain panic disorder patients at a constant level of medication, if no problems emerge, than to subject them to this cycle of cessation and recommencement of medication – regardless of the class of medication in question.

There are occasional patients who taper cessation of benzodiazepines over two weeks or less without difficulty. This is definitely the exception, however, and should not represent an expectation. Such a rapid taper can be dangerous and even life-threatening.

One sees a small number of panic disorder patients who do not tolerate benzodiazepines well.

It is widely regarded as canonical that benzodiazepines cannot be used in the treatment of patients with a history of alcohol abuse. However, with proper supervision, this is not necessarily true. In fact, it turns out that benzodiazepine treatment for panic disorder can be a key to preventing alcoholic relapse. Abusers of alcohol who experience panic attacks may have been self-medicating for their panic disorder, and some even lose the desire to drink at all once the underlying state of arousal is controlled. This seems particularly true in the case of measured drinkers, those who drink regular amounts at regular times of day.

Both anecdotal reports and published research emphasize that the definition and manifestations of tolerance, dependence, and addiction are commonly misunderstood among physicians and even among a surprising number of psychiatrists. Experts in the use of benzodiazepines report that tolerance normally develops to their sedative, cognitive, emotional, and motor effects, but not to the autonomically anxiolytic effects, i.e. not to their ability to interrupt the neurological "panic circuit." Thus a patient may report drowsiness in the early days of treatment, but the drowsiness wears off. Once an effective dosage is found in the first weeks of treatment, this dosage (or, after a while, a lower one) will eliminate panic attacks and most dysautonomia for years and decades. Patients on such a maintenance dosage should be regarded as "medically dependent," just as many patients are dependent on glucophage or insulin or anti-hypertensive medications that exhibit a rebound effect on cessation – or on Paxil, for that matter – but should not be regarded as "addicted" in the absence of all the other criteria that characterize addiction.

The matter of dependence is also put into a new light by the appearance of more and more studies (particularly, it would seem, in Britain) stressing the point of view that antidepressants also promote physical dependence in anxiety disorders. A World Health Organization report goes so far as to state plainly that SSRI withdrawal is a more difficult problem today than Valium withdrawal was in its time. Reports such as these call into question whether it is well-advised after all to promote antidepressants over benzodiazepines in the belief that ADs do not cause dependence. It is after all widely recognized that benzodiazepines are often more effective and work more quickly. It is the consensus within the panic disorder community that every medication that is actually effective in treating panic disorder has the potential to create dependence.

Dr. Robert DuPont, a former director of the [United States] National Institute On Drug Abuse, has stated that "the concept of using the lowest possible dose of a benzodiazepine for the shortest period of time is inconsistent with sound clinical practice. This concept needs to include the important qualification that our goal is to maximize the patient's ability to function well and to enjoy life. Benzodiazepines are among the safest and most effective treatments in all of medicine, including their role in the treatment of panic disorder." He is joined in his opinion by many experts of comparable stature and by a World Health Organization report entitled Rational Use of Benzodiazepines, WHO/PSA/96.11. (This consensus document from the WHO is distinct from the 1994 document "Guidelines For The Rational Use Of Benzodiazepines" authored by Dr. Heather Ashton.) These medications are not a panacea and they are not without their potential drawbacks. However, they have been the subject of such a fierce backlash from their extravagant misuse and overprescription in the 1970s (similar to but exceeded by the SSRI bonanza of today), and have been the target of such unscrupulous propaganda from the promoters of other drugs which still, unlike most benzodiazepines, enjoy the profitability of patent protection, that most doctors are unfamiliar with the actual evidence that should govern their prescribing. The evidence clearly shows that at present there is no other class of medication that works nearly as well, in nearly as large a proportion of the panic disorder population, as benzodiazepines; it shows that dependence is real but is not unique to benzodiazepines and thus not a major differentiating factor; and it shows that there is no other population less prone to addiction or abuse of benzodiazepines prescribed as part of a comprehensive management strategy.

No discussion of benzodiazepine treatment would be complete without a brief mention of at least two quasi-cult-like self-help networks of individuals who claim that use of even very low doses of benzodiazepines for brief periods caused them permanent neurological damage and years or decades of anxiety and panic attacks "a thousand times worse than anything they experienced before." (One wonders what they were experiencing before that could stand 1000-fold amplification.) With a vehemence and personally directed vitriol strongly suggestive of fanaticism, they reject any suggestion that they might instead be suffering from untreated anxiety or panic disorders, since the symptoms they describe when one pins them down to specifics sound very much like the symptoms of uncontrolled, advanced panic disorder.

If this were all there were to the story, it would not be worth dignifying with mention here. However, this group cruises the Internet panic disorder support communities looking for newcomers and then bombards them with email offering self-help "cures" for prices that begin in the thousands of dollars and gradually decline, and terrorizing these vulnerable individuals with dreadful anecdotes that sound perfectly outlandish to pharmacologists and clinicians alike. Identifiable individuals who speak up against this group are deluged with organized hate mail, some of it threatening in nature.

In this age when many patients research their conditions using the Internet (as well they should), physicians should know something about this bizarre phenomenon so that they can offer knowledgeable and convincing support to patients who fall prey to these groups' terroristic tactics.

Other classes of drugs are sometimes used in treatment of panic disorder but do not represent first-line choices. Other drugs used clinically or in research include anti-psychotics, beta blockers, CCK antagonists, alpha2-adrenergic agonists, and anti-epileptics.

Although the author lacks statistical data, there are a notable number of reports from patients experiencing good results with Effexor XR (which seems for unknown reasons to behave very differently from its non-extended-release version Effexor), often after having bad results with SSRIs or simply having run the gamut of the major SSRIs, only to have each one in turn "poop out" (if the reader will excuse the scientific jargon) after a period that typically begins at a number of years but tends to shorten with each successive replacement drug.

Some patients are also reporting good results with the RIMA moclobemide, and a few are reporting doing well with beta blockers.

Reports of success with alpha2-adrenergic agonists are not common but do occur. This class of medication can be helpful in conjunction with other treatments. The same observations apply to beta blockers.

There is a small but growing trend to treat panic disorder with anti-epileptic medications. Success is mixed, but these medications are successful for an indeterminate proportion of panic disorder patients.

There has been a puzzling surge of reports over the last year of doctors resorting to antipsychotic medications in the treatment of panic disorder, often before even trying all of the mainstay classes of medications, and often with unfortunate consequences. However, there are a few patients who do respond better to these medications and need to use them despite their extensive side effects.

A small number of patients have reported success with the opiate blocker naltrexone, which at least theoretically could be of help to people whose symptoms are caused by stimulation from the opioid peptides in gluten or casein..

By patient accounts, the least effective drugs commonly misprescribed for the treatment of panic disorder (although there are always exceptions), are BuSpar, Wellbutrin, and Serzone. Serzone has the particular complication, of which an alarming number of prescribing physicians seem unaware, of substantially inhibiting the metabolism of other psychotropic drugs, thus making it difficult to calibrate effective dosages. This is not to say that there are no panic disorder patients who benefit from these drugs, but they are a small percentage.

There is a trend toward polypharmacy in the treatment of panic disorder, with drugs added to treat the side effects of drugs added to treat other side effects. Often drugs are piled on in an attempt to avoid or make acceptable the obvious but in some circles politically incorrect choice of benzodiazepines, or in blind adherence to the doctrine that SSRI medications simply must work for all. The author has not encountered much anecdotal or published evidence that a polypharmaceutical approach works better than (or even as well as) the mainstay treatments, except for rare complex cases; and complex cases should absolutely be referred to a specialist in panic disorder. Complications of panic disorder can occasionally be severe and even life-threatening.

Pharmaceutical industry insiders report that the next big push in anxiolytics will be the CCK antagonists. Already marketed successfully outside the United States for this purpose (as well as for treatment of GERD and IBS), these are reportedly intended for an aggressive push into the American market when a favorable point is reached in the product life cycle of the SSRI medications. (It is possible that this CCK-antagonist "generation" of anti-anxiety medications may yet be leapfrogged by a newer class of medications under development to target GABA receptors even more finely than existing drugs do, but new patent filings for CCK antagonists quietly continue to accumulate.) For the time being, the SSRIs are still among the most profitable of pharmaceutical products in the United States and are among the least vulnerable to competition. Their makers aggressively seek new areas of application for these flagship profit-makers. But factors already at work to shift the market equilibrium include recently published major studies identifying the chromosomal locus and the importance in panic disorder, GERD, and IBS of defects in CCK metabolism, increasingly general awareness of SSRI complications, and the proliferation of lawsuits alleging improprieties in the conduct of clinical trials for SSRIs and the withholding of unfavorable research data. It is simply a matter of time before we see the next shift in focus.

It is evident to the author that at least in the case of newer psychotropic drugs, one should trust postmarketing research and the reports of experienced clinicians over the published results of Food and Drug Administration -supervised clinical trials for approval of new drugs.

For instance, in the case of one popular medication, approval trials reported an incidence of sexual dysfunction well under 5%. Yet today experienced clinicians are in agreement that over the same length of time covered by the original trials, the true incidence of sexual dysfunction is at least 60%. Some claim up to an 85% incidence in their practices. And such numbers have been substantiated in postmarketing research. How does one explain such discrepancies in the light of FDA oversight?

Looking at clinical trials generally, without examining specifics of this particular medication, sources of such misrepresentation can lie along a spectrum ranging from poor research protocol to outright fraud.

In the area of protocol, test subjects for the medication above were not asked about sexual dysfunction and felt reluctant to bring it up, especially given the relatively short duration of the trial. And while randomization may have been scrupulous within the trial, it has come to light that all candidates for the study, whether they ended up on placebo or the experimental drug, were carefully hand-picked for previously demonstrated tolerance of that drug's side effects.

(As an aside, it is interesting to note that physicians participating in clinical trials state that the concept of double-blind placebo control is little more than a convenient fiction in the testing of psychotropic medications, and a somewhat ridiculous fiction at that. They say that in most cases, the high incidence of dramatic side effects with psychotropics makes it perfectly obvious to the health care personnel, if not indeed to the patients, who is taking the medication and who is taking the placebo. For this reason, drug-against-drug trials are often more useful.)

Looking again at the general case, there is also the question of fraud. Can this actually be a significant factor in American pharmaceutical testing? Doesn't the FDA ensure the validity and integrity of such testing? The General Accounting Office seems to think otherwise. The New York Times, in a 1999 series of articles on fraud in drug approval trials, cited a GAO report criticizing the FDA for repeatedly failing to investigate whistle-blower reports of fraud and for colluding in some cases to cover up evidence of its commission. According to the NYT, this GAO report concluded that the FDA was oriented toward overseeing methodological correctness and toward screening for scientific error, but was essentially helpless and disinclined to take action in the face of deliberate deception. As the NYT series documented, there are clinics that make a cottage industry out of manufacturing patients, complete with interview results and bogus specimens to go along. Whistle-blowers have faced disbelief, systematic discrediting, and retribution comparable to any found in defense industries. Far from ferreting out and assuring prosecution of such fraud, certain FDA employees and their supervisors seem to have facilitated its commission and cover-up, according to the GAO. This may be one reason that American pharmaceutical testing is viewed with skepticism in for instance Germany and Finland, two of the thirty-odd countries consistently recognized by the World Health Organization (and other international observers) as practicing higher standards of health care and governmental oversight than the United States.

The FDA is just another government agency. While it is held in peculiar veneration by the medical community, there is no reason to expect it to be any more competent or any more scrupulous in its regulatory and inspection/examination activities than other agencies such as the Internal Revenue Service, the Department of Energy, or the Department of Agriculture. (Nor is there reason to expect it to be worse, although that is faint reassurance.) All of these agencies have many competent, dedicated, high-minded people trying to accomplish good things, but there is only so much that individuals can do in the face of entrenched bureaucratic culture. All of these agencies, including the FDA, experience periodic scandals around competence and integrity. The FDA comes under blistering criticism from European medical authorities (not to mention domestic sources) for its pattern of approving medications for which pre-approval research demonstrated serious problems, only to end up withdrawing approval under pressure from consumer organizations after unacceptable levels of fatality or injury appear. It is naive to exaggerate the competence or integrity of the United States' drug approval and oversight compared to taxpayer service, oversight of the nuclear energy and weapons industries, or inspection of meat and agricultural products. Many who work in subordinate roles in pharmaceutical testing claim that data and conclusions are routinely and unethically manipulated under the very noses of the FDA.

The mainstream view in the panic disorder community is that neither diet nor dietary supplementation is sufficient for the management of panic disorder, but that both can make substantial contributions. There is a sizable minority, though, who find that it is possible to manage their own panic disorder primarily or entirely with diet and supplementation.

The four most dramatic examples of dietary management are elimination of caffeine-like stimulants, adoption of a gluten-free/casein-free diet, supplementation with the amino acid taurine, and adoption of a ketogenic diet. Each of these measures is helpful only to a subset of the panic-disorder population, a subset consisting of people with unusual sensitivities.

Elimination of caffeine-like stimulants. Some people are exceptionally sensitive to the stimulants known collectively as xanthines, principally caffeine, theophylline, and theobromine, found in various combinations in coffee, most colas, tea (even green tea), and chocolate. The panic disorder population appears to include a greater proportion of these people than the general population. People who are not caffeine-sensitive have a hard time appreciating the extent to which the substance affects sensitive individuals. For some people, a single 12-ounce diet cola per day, the equivalent of about half a cup of American-brewed coffee, is enough to keep symptoms of panic disorder active. Quite a few people have reported that a single cup of decaffeinated green tea daily was the difference between experiencing heart palpitations and not experiencing them.

Caffeine must be discontinued carefully. People seem to have the best success by tapering gradually – ounce by ounce at the end, if necessary – before completely quitting. Many people take an NSAID like ibuprofen or aspirin for a few days starting the day before complete elimination in order to avoid the ferocious caffeine-withdrawal headaches that can otherwise result.

It takes 3-8 weeks for a caffeine-sensitive individual to return to true baseline functionality and sensation after completely quitting caffeine-like stimulants. Prominent symptoms during the withdrawal phase are likely to include headache, depression, loss of energy and initiative, muscular stiffness, constipation, temporary changes in vision, reduction in coordination, minor edema, sleeping several hours longer than usual, and markedly reduced athletic, physical, and mental performance. Because of the way caffeine affects biochemical production of energy to power muscle cells and brain cells, one is quite literally not as strong and not as smart during caffeine withdrawal, but full function does return after withdrawal is complete. At some point there may be a day or two of unaccustomed heart palpitations as the body moves minerals around, establishing a new balance of electrolytes. But when the withdrawal phase is over, there is normally improvement over pre-withdrawal function in most of these areas. Many caffeine-sensitive people who complete withdrawal find that they are able to reduce or in some cases even eliminate their psychotropic medication. But this should not be an expectation; the reason to try caffeine-elimination and persist until withdrawal symptoms disappear is simply to feel better, function better, and quite likely lessen some of the dysautonomia.

Gluten-free casein-free (GF-CF) diet. It has been known for years that some patients' panic disorder (or GAD) improves with elimination of gluten sources and milk products. Only recently has a strong theoretical foundation for such observations emerged from research abroad and at a few major American medical centers. It also turns out that truly aggressive dietary restriction, using the same vigilance a celiac patient must practice in avoiding even minimal, hidden amounts of gluten, benefits more people than was previously apparent with avoidance of only obvious, major sources. Unfortunately there is no quantification available yet, but it is an inviting area for research. Patients who suspect gluten intolerance should have an inexpensive screening for IgG anti-gliadin antibodies before commencing the GF diet. With strict avoidance, the antibodies disappear in 3-12 months. However, one is still intolerant to gluten. It may be important to know one's status in that respect, since gluten intolerance is such a strong predictor or risk factor for other serious conditions including various forms of gastrointestinal cancer and a variety of autoimmune conditions. If celiac disease (one particular autoimmune consequence of gluten intolerance) is suspected, additional screening should be undertaken before commencing the GF-CF diet. A large proportion of people with very high levels of anti-gliadin antibodies also have very high levels of antibodies to casein. Even without such high antibody levels, the gluten-intolerant patient is likely to experience psychotropic effects from casein, although the top American experts in celiac disease note that some patients recover the ability to tolerate milk products as their digestive tract regenerates itself.

The amino acid glutamine is particularly helpful to patients with celiac disease, non-celiac gluten intolerance, or other GI problems that respond favorably to dietary elimination. The digestive tract is one of the fastest-rebuilding parts of the human body and one of its principal building blocks is glutamine. Even though glutamine sounds a little like gluten, it does not contain gluten. When pure it is a safe nutrient and actually an important nutrient for gluten-intolerant individuals.

Taurine supplementation. Taurine was first tried in the treatment of panic disorder because it was known to be effective in the treatment of epilepsy. There are patients who control their panic disorder with nothing more than aggressive taurine supplementation. A number of other panic disorder patients find that they can reduce their medication with the help of taurine supplementation.

Ketogenic diet. The overall healthfulness of a ketogenic diet is certainly debatable. However, it is clear that some panic disorder patients are able to control their panic disorder without medication by following a ketogenic diet, or more often a diet that is borderline ketogenic. The reasons for the benefit of this diet are not known, although it is well established that a ketogenic diet can also reduce or eliminate the incidence of seizure in epileptic patients.

Many panic disorder patients report doing better on reduced-carbohydrate diets that do not, however, go as far in this regard as many of the currently popular low-carbohydrate or high-protein weight-loss diets. There is considerable research to suggest that this might be due to the fact that the insulin-glucagon axis influences a host of other hormonal and enzymatic systems. However, there has been little direct study of carbohydrate restriction in the treatment of panic disorder, and the only connection that seems obvious and likely to be uncontroversial is the adrenaline response that can be provoked by falling blood sugar levels. Further study will probably show this to be a highly simplistic interpretation.

Magnesium supplementation. Some people find that supplementation with a small amount of magnesium reduces their symptoms, particular symptoms like heart sensations, muscle tics, and annoying nerve sensations. Too much magnesium will cause diarrhea and exacerbate the body's imbalance of minerals, but just enough can noticeably improve nervous function in people who have a deficiency of magnesium in nerve cells. (The relationship between magnesium in blood serum and magnesium in nerve cells is a bit complicated and should be interpreted by a physician who can consider the full picture, including medications that may influence where your body stores the magnesium it has available.)

Fat restriction, supplementation with essential fatty acids. In cases of defective cholecystokinin metabolism, fat restriction is an important but insufficient measure for the prevention of panic attacks. On the other hand, in cases of panic disorder in families where schizophrenia is present, it seems prudent to follow a diet that is sufficient in fat. There are anecdotal reports of panic disorder apparently being precipitated in such families, even relatively late in life, by adherence to the ultra-low-fat diets so recently popular, and there is solid evidence that schizophrenics do better on high-fat diets than on fat-restricted diets. While the author has uncovered no proof that a diet relevant to the care of schizophrenia is also relevant to the care of panic disorder in families where schizophrenia is present, it seems prudent, at this point in the understanding of such panic disorder and of schizophrenia, for panic disorder patients from schizophrenic families to follow the dietary measures that should be followed in schizophrenia. These include strict avoidance of casein (a milk protein) and gluten and gliadin (a family of proteins found in wheat, oats, barley, rye, and some exotic grains), and the inclusion of sufficient amounts of fat, perhaps above-average amounts of fat. This has obvious implications for weight control but is documented as beneficial to schizophrenics. Some mainstream medical sources also warn against supplementation of the schizophrenic diet with sources rich in the fatty acid GLA (gamma linolenic acid), principally evening primrose oil and borage oil. However, where the author has observed such warnings, they have occurred in contexts attempting to "debunk" the purported benefits of these supplements, and therefore the standard of proof was probably lower than if the authorities in question had been asserting a benefit rather than a danger. (In other words, there is in medicine, as opposed to the body of science proper, one standard of proof applied to a claim like "substance X may cause hair to grow back on your balding scalp," but quite a lower standard of proof applied to a scare-claim like "not only will substance X not cause hair to grow back on a balding scalp, but furthermore, substance X may cause the hair you still have to fall out, and maybe your teeth, too.") Nonetheless, prudence should prevail. There is solid evidence that GLA is contraindicated in patients with seizure disorders, and it may be wise for the panic disorder patient from a schizophrenic family (or an epileptic one) to avoid GLA-rich supplements.

Other dietary measures. It is beyond the scope of this document to treat the full complement of dietary supplements being used in the treatment of panic disorder or investigated for such use. There are simply too many to discuss, and the results are too individual to each patient. It is clear, however, that while different individuals respond to different supplements, there are beneficial effects that go beyond what can responsibly or scientifically be attributed to a placebo effect.

The supplements that prove effective are not the obvious choices – for instance, only a very small number of people report significant benefits from the well-known anti-anxiety supplements like Kava kava or Passionflower. Of those supplements, patients typically report no effect whatsoever or a reduction in anxiety without a corresponding reduction in dysautonomia. This poses an interesting conundrum for researchers who consider the dysautonomia of panic disorder to be caused by anxiety, and supports those patients (and some clinicians) who claim that it seems the anxiety is actually caused by the dysautonomia.

Nor does the antidepressant supplement St. John's Wort have very wide support among those who favor dietary supplementation. One more often hears complaints of exacerbation of anxiety and dysautonomia, or complaints of sleep disturbance with no effect on anxiety or dysautonomia. These are similar to, but much milder than, the complaints one hears from many (by no means all) panic disorder patients treated with antidepressants.

Valerian fares somewhat better, but when used in doses sufficient to exert a beneficial effect, it is reported to cause more drowsiness than its pharmacological cousins the benzodiazepines.

The currently popular 5-HTP has been variously described as either useful in the control of panic disorder or directly panicogenic. (Of course the same observation applies to 5-HTP's very distant cousins, the SSRI medications.) 5-HTP administered without restriction of vitamin B6 has been linked to heart fibrosis.

Very recent double-blind, placebo-controlled research suggests that Gotu Kola is somewhat effective against the variant of panic disorder arising from the genetic aberration in cholecystokinin metabolism. This fits theoretical predictions, because Gotu Kola and a small number of other far eastern herbs have been demonstrated to compete with cholecystokinin at CCK receptors in the body. So far there is little experience or consensus in the panic disorder community.

Supplements more commonly reported to be mildly beneficial are those that support parasympathetic nervous system function (B-vitamins, alpha-lipoic acid, precursors or cofactors for the synthesis of acetylcholine and nitric oxide), supplements promoting cardiac stability (Q10, carnitine), supplements soothing the digestive tract (chamomile, bismuth, glutamine, MSM, occasionally peppermint - although the last can also be counterproductive)… It is simply not possible to do more than scratch the surface lightly in this discussion.

There is no clear consensus and there are almost no formal studies of the effects of special diets or supplementation on panic disorder. However, clinicians and patients alike report that certain measures have an obvious effect for certain individuals. One might suppose that panic disorder patients would be more susceptible than most to a placebo effect; but in fact, panic disorder is so difficult to control without medication, and panic disorder patients are so reluctant to resign themselves to a lifetime of medication, that they often try many dietary supplements in a systematic fashion, generally rejecting the preponderance of them as ineffective. For a panic disorder patient to report that a supplement is beneficial, even if it is not sufficient to fully control the condition, is an exceptional event that should usually be taken seriously – especially if it makes possible a lighter use of medication. One cannot easily feign control of panic disorder or delude oneself about doing so, and there is little incentive to try.

Panic disorders are a family of disorders the understanding of which is still in its infancy. Much of what we thought we knew only a few years ago has been proven false. (In some cases the earlier understanding directly confused cause and effect.) There is far more that is not known about panic disorder than there is that is known; however, a nascent body of research in several new directions is revealing some of the secrets.

Panic disorder is not a monolithic condition. Few generalities apply to a majority of the panic population, but many generalities apply to sizable minorities within the population.

Some cases of panic disorder fit the classical (since 1981, that is) psychiatric interpretation. Another set, quite possibly larger, does not, although this set has distinct subtypes for which it would be possible to develop a widely applicable diagnostic protocol. Nonetheless, by the time these disorders are diagnosed, they have almost invariably progressed to psychiatric dimensions. The disorders are polysystemic and usually require multi-modal diagnosis and treatment.

Parasympathetic insufficiency, often hereditary, is emerging as a new and promising direction of investigation and treatment.

Problems in CCK metabolism have been linked to panic disorder by multiple independent researchers for some time now, but 1999 saw publication of a major study that claims to have identified the genetic defect in question.

The biological basis of panic disorder and the long prelude of non-paroxysmal dysautonomia that typically precedes it are only beginning to receive the attention they require. It will be a tremendous step forward when there is general recognition that treatment of panic disorder must aim not only at total elimination of "panic attacks" (or dysautonomic paroxysms) but at elimination of the associated chronic dysautonomia as well. Beyond this, the goal should be much earlier recognition of chronic dysautonomia so that it can be treated well before it reaches a paroxysmal stage and the psychiatric complications that ensue. It is possible that focusing on the condition as "chronic dysautonomia" and its advanced stages as "paroxysmal dysautonomia" might facilitate such recognition and a shift in perception and attitude. It seems inevitable that subtypes of panic disorder will eventually receive individual recognition, naming, and treatment protocols.

Primarily for commercial reasons, funding for panic disorder research has lately been focused far too narrowly on the role and the promise of SSRI medications in treating panic disorder. This shift in focus to follow the availability of research funding interrupted a number of other fruitful areas of earlier research. Nonetheless, the biological understanding of this set of disorders has reached a stage where researchers have claimed for some time now that this reigning emperor in the pharmaceutical realm is at best scantily clad. Apparently the emperor's successor, in the form of CCK antagonists, has already been designated and is awaiting coronation, but only time and experience will tell whether the next panacea is more universally effective than the current one. The author is extremely doubtful that any panacea will ever be found for what we know today as panic disorder – it is too clear that distinct and distinguishable disorders are being carelessly lumped together as one.

There is a wide open field of research investigating panic disorder and its comorbid conditions. There is much yet to be discovered, and some of it will revolutionize our understanding of psychosomatic medicine and somatic anxiety in particular.

Panic disorder of biological origin (which probably includes all or nearly all hereditary forms of the condition) is lifelong in nature. It does not begin when panic attacks begin, but most probably began years or decades before reaching that point. Remissions may occur, but relapse should be expected. This is true regardless of the method used in treating the condition – it is a myth that antidepressants provide a "cure," except in a limited subset of cases where panic disorder is secondary to another condition.

Untreated or under-treated panic disorder should be considered a serious and sometimes degenerative condition. It puts significant strain on a number of body systems, and the goal of treatment should be not merely the elimination of panic attacks but the elimination of the underlying or associated autonomic hyperarousal as well.

Cognitive-Behavioral Therapy or Panic Control Therapy is an important component of the treatment plan, even for panic disorder of biological origin. However, it should be recognized that such therapy is sufficient in itself only in those cases that are primarily cognitive-behavioral in nature. Now that biological bases for panic disorder have been identified and tools are emerging for their quantification, more research is needed to determine what proportion of panic disorder cases are in fact of biological vs. psychological origin. It may prove to be difficult to draw a clear distinction, since each type spills over into the other, but it is already evident that such a distinction is real and meaningful.

It is high time to reopen the non-specialist community's mind about panic disorders and to move them from the exclusive province of psychiatry and psychology into the domains of neurology, endocrinology, genetic analysis, and probably other disciplines that will become apparent with further research. This is exactly what growing numbers of researchers and practitioners are doing.

 [ Top ] [ Index ]

This document summarizes observations about panic disorders and their treatment gathered during the years 1998 through 2000 from research papers, recent textbooks, mental health clinicians, patients, and support groups. It is not intended to provide medical advice. Nor is it intended to advocate original points of view in the scientific or medical arenas. Rather it is intended to integrate and disseminate a variety of emerging insights often omitted from the narrow classical view of these disorders.

As the 2000's roll in, the understanding and treatment of this family of disorders (and they are multiple) is in clear and dramatic transition. Ongoing research about modalities of diagnosis and treatment is beginning to bear fruit that will profoundly influence both the medical and psychological perspectives on these conditions. The viewpoints advanced in this paper are for the most part not of the author's own invention – they represent viewpoints that have arisen and show growing support in the researcher, clinician, and patient communities. In one sense, they represent a consensus; but in another sense, the term "consensus" is misleading because the field is still in turmoil, with different schools of thought pursuing different avenues of research and fighting to advance different agendas.

This paper represents the author's good-faith efforts to gather the points of view that appear, in his lay judgment as a hardheaded observer and a patient himself, to have the best foundation and most correspondence to the reality that he and countless other patients live every day. Sources, which are not cited, include published research papers (most available through MedScape or MedLine), published textbooks and trade books, articles in WebMD and the New York Times, the opinions of clinicians specialized or experienced in the treatment of panic disorders, and the experiences of some hundreds of individuals who have shown up in various Internet-based support communities focused on panic disorders or other anxiety disorders. Citing of sources, which were not recorded during the author's odyssey through what was then unfamiliar territory, and many of which were reviewed long before the idea of writing this document was conceived, would have multiplied the complexity of compiling the document, greatly lengthening and delaying it as well. Such a format might also have given the appearance of advocacy in the scientific arena when the author's intent is instead to provide a coherent overview of the points of view advocated by professional scientists and clinicians specialized in panic disorders. Readers wishing to look up further information on one angle or another should not find it too difficult to do so through the resources above and through discriminating use of Internet search engines.

Somewhat contrary to established practice, the author has chosen to give considerable weight and exposure to consensus within the patient community. It is this community that lives the day-to-day reality of panic disorders, this community that feels the effects and the side effects of medications and therapies every day and knows the reality and nuance of our experience. Our doctors and therapists can (and certainly do!) try to tell us what we are and are not experiencing based on the most interesting presentation they attended at a recent conference, or on the glowing reports of a new medication (this one miraculously free of all the side effects that have come to be associated with each of its predecessors), or on the doctrines that have brought recognition and prestige to their alma maters; but we as patients have only one agenda, which is a return to well-being and normality; and we have only one truth, which is the experience within our own bodies that most of our doctors and therapists will never (to their great good fortune) be able to do more than imagine.

(Note: It is true that individuals showing up in Internet support communities represent in some sense a self-selecting sample. However, the same observation could be made with equal or greater validity about the individuals who choose to remain in treatment with a particular clinician or about those who choose not to drop out of controlled clinical trials. Indeed even a telephone book represents a self-selecting sample. Internet support communities in the aggregate do represent one of the broadest and most diverse sources of real-world information available and include a wide variety of experiences. In the case of panic disorder, these groups include individuals who are well-controlled and those who are not; those who claim to be "cured" or in remission without ongoing treatment and those who do not; those who favor one class of treatment and those who favor another; those who agree with their doctors' assessment of successful or unsuccessful treatment and those who make a different assessment. Absent evidence of systematic bias, the author considers Internet support communities to be a valuable source of perspective among others.)

Understanding of panic disorders is still in its infancy. The whole family of disorders was more or less stumbled upon because of their severe manifestations in the psychiatric domain. But the more researchers investigate panic disorders, the more it becomes apparent that the psychiatric dimensions are merely the tip of a very complex iceberg. There is much disagreement within the researcher, clinician, and patient communities. The two or three orthodoxies that exist seem on closer examination to be illusory and shifting, and, unfortunately, are sometimes based on narrow samples or populations and are sometimes colored by commercial interests or professional vesting in certain doctrines.

Ultimately the viewpoints and conclusions presented here, which are meant to represent some diversity of opinion, are of the author's own selection and are not necessarily endorsed by any other individuals or institutions except where specifically attributed. So in one sense this essay is merely one person's collection of opinions. But this collection has been reviewed by researchers, clinicians, and patients alike; and while no essay could please all readers, and most especially not those who are vested in one particular point of view to the exclusion of others, the author feels reasonably justified in stating that in another sense, this document is not only built largely on published research and teaching materials, but also does meet with general approval and endorsement from at least a substantial segment of the researcher, clinician, and patient communities.

The online source for this document resides at http://www.panic411.org/

This document may be linked to or printed for limited noncommercial distribution by individuals. Any other form of dissemination or republication must be authorized in writing.