What causes panic? - Panic disorder

Anxiety: A Very Short Introduction - Daniel Freeman, Jason Freeman 2012

What causes panic?
Panic disorder

In 1959, the US psychiatrist Donald Klein administered imipramine, a recently developed antidepressant, to a number of patients who were probably suffering from panic disorder (though this was not a term in use in those days). Klein was not optimistic about this new treatment:

it was more a case of our not knowing what else to do for them, and thinking that perhaps this strange new safe agent with peculiar tranquilizing powers might work.

But imipramine seemed to produce some remarkable changes in Klein’s patients: within a few days, their panic attacks had disappeared. However, their general anxiety levels remained essentially unchanged. This led Klein to make a ground-breaking distinction between panic and anxiety. Panic, he argued, was a phenomenon in its own right, and not — as was then believed — simply a feature of anxiety. In so doing, Klein paved the way for the identification of panic disorder. (He also discovered that agoraphobia was rare in people who had never experienced a panic attack — another major breakthrough.)

Klein continued to investigate panic, and found that he could provoke it by dosing individuals with substances such as sodium lactate, or asking them to inhale carbon dioxide. Panic, he concluded, was a product of biological processes, and specifically a reaction to a perceived lack of breathable oxygen. (Levels of lactate in the brain rise when we cannot breathe properly: for example, if there is too much carbon dioxide, and too little oxygen, in the air.)

We propose that many spontaneous panics occur when the brain’s suffocation monitor erroneously signals a lack of useful air, thereby maladaptively triggering an evolved suffocation alarm system.

Fear of suffocation features in some panics. But over the years much of the evidence for Klein’s theories has been questioned. No one, for example, has been able to identify where exactly in the body the suffocation alarm system is located. Moreover, scientists have noted that panic can be induced by a vast range of substances operating on differing aspects of human physiology — which implies that no single biological mechanism is involved. Similarly, several classes of medication can block panic attacks. And approximately 50% of people given sodium lactate, for example, do not respond by panicking: it’s not an inevitable response.

Convincing demonstrations that panic attacks have a psychological component present the most serious challenge to Klein’s account. For example, Ron Rapee and colleagues asked a group of people suffering from panic attacks, and another with social phobia, to inhale a mixture of carbon dioxide and oxygen. Half of the participants were not told what to expect from the inhalation; the other half were warned that they might experience sensations associated with a panic attack (for example, chest tightness, breathlessness, dizziness).

Whether or not they received an explanation made no difference to the participants with social phobia: they reacted to the gas in exactly the same way. But things were very different for the participants who suffered from panic attacks. Those who hadn’t been told what to expect were much more likely to panic than those who were. Similar studies have shown that, if you simply give participants an illusion of control over the experiment — for example, by leading them to believe (erroneously) that they can reduce the supply of whichever substance is being administered — they are much less likely to panic.

Experiments like these have been crucial in the development of a psychological account of panic, which attaches primary importance not to physiological processes, but to thoughts. The work of David Clark has been hugely influential here. Clark argues that:

Panic attacks result from the catastrophic misinterpretation of certain bodily sensations.


7. David Clark is a world leader in the understanding and treatment of anxiety disorders, and a pioneer of new cognitive therapies that have transformed clinical outcomes for panic disorder, social phobia, and (with his wife Professor Anke Ehlers) post-traumatic stress disorder. Chair of Psychology at Oxford University, and Director of the Centre for Anxiety Disorders and Trauma at the Maudsley Hospital, Clark has also been a crucial figure in the UK Health Service’s remarkable Improving Access to Psychological Therapies scheme

So carbon dioxide or sodium lactate or any of the myriad alternatives used in laboratory studies don’t cause panic, at least not directly. What produces panic is the ’catastrophic misinterpretation’ of the physical effects produced by that substance.

Let’s illustrate Clark’s theory with a fictional case study:

Martin experienced his first panic attack a couple of years back. The attack seemed to Martin to have come out of the blue, though later he realized that he’d been under exceptional stress for a number of weeks because his father had been seriously ill. During that first panic attack, Martin had been convinced that he was dying of a heart attack. Although he was assured by his doctor that his heart was perfectly healthy, in the weeks and months that followed, Martin was constantly on the look-out for what he imagined were the signs of an imminent cardiac arrest: for example, his heart seeming to beat faster than normal, chest pain, or difficulty breathing. When he detected these physical changes, he became so anxious that a panic attack often followed.

In time, Martin became less concerned about the prospect of a heart attack; instead, his fear centred on the prospect of experiencing another highly distressing and unpleasant panic. In particular, he dreaded losing control of himself during a panic attack. And so Martin monitored his physical sensations just as keenly as ever; all that changed was his view of what they meant (a panic attack rather than cardiac arrest). Martin became reluctant to leave his home in case he ’humiliated’ himself by experiencing a panic attack in public.

Martin’s panic attacks, then, were initially triggered by mistaking normal signs of stress or anxiety for symptoms of imminent collapse and perhaps death. Later, Martin viewed these normal sensations as indications of an imminent disastrous panic attack — a misinterpretation that, ironically, helped produce the very event he feared. Incidentally, it’s not only the bodily changes produced by anxiety that can trigger panic: any apparently odd sensation can do it — for example, when falling asleep, or taking physical exercise, or after having drunk too much coffee.

Like many people with panic disorder, Martin was hypersensitive to bodily changes. The psychologists Anke Ehlers and Peter Breuer demonstrated this tendency when they asked a range of volunteers, including 120 individuals with panic disorder, to silently count their heartbeat without taking their pulse. What they found was that those with panic disorder were much more accurate in their estimates than the other participants. Subsequent research has found that the same is true for children with panic problems.

Martin’s reluctance to leave his home would probably be diagnosed as agoraphobia. His avoidance is actually a safety behaviour. Such behaviours figure prominently in most cases of panic disorder, just as they do in the other anxiety disorders. By staying at home, Martin denied himself the opportunity to learn that he could experience the symptoms of a panic attack without anything disastrous occurring, be it a heart attack or losing control.

Support for the catastrophic interpretation theory also comes from research into the predictors for panic attacks — that is, the factors that make people vulnerable to panic. A few years back, the psychologists Norman Schmidt, Darin Lerew, and Robert Jackson were given access to more than a thousand young people undergoing five weeks of basic military training at the United States Air Force Academy. The training is extremely tough — recruits endure punishing physical work-outs on limited sleep, are subjected to constant observation and evaluation, and, in many cases, are away from family and friends for the first time in their lives. Because of the high level of stress, it’s perhaps hardly surprising that some recruits develop problems with anxiety — and with panic.

But Schmidt and colleagues found that susceptibility to panic attacks wasn’t random. On the contrary, it could be predicted by looking back to how the recruits had scored on a specific psychological test they’d taken at the very start of training. This test measured anxiety sensitivity, which is the extent to which a person believes that the physical sensations experienced when excited or anxious (increased heart rate or breathlessness, for instance) are harmful. The higher a recruit’s anxiety sensitivity rating, the more likely it was that they’d gone on to experience a panic attack.

Research like this has led some psychologists to see anxiety sensitivity as a risk factor for panic disorder. It certainly fits well with the psychological accounts of panic we’ve just explored: anxiety sensitivity can be seen as an underlying tendency towards the kind of ’catastrophic misinterpretation of bodily sensations’ that is believed to drive panic disorder.

That tendency may be fostered by the attitudes to panic and illness we are brought up with. Research has shown that people prone to panic attacks are more likely to have witnessed their parents being highly anxious or panicking, and then misinterpreting those panics as a sign of illness (for example, by taking it especially easy or asking for special attention). They are also more likely to have lived with people suffering from chronic physical illnesses; this may make them especially sensitive to the symptoms of sickness.

If life becomes especially stressful, this vulnerability may lead to panic attacks and perhaps even panic disorder. On the other hand, it’s possible that anxiety sensitivity measures not so much a vulnerability to future panic but the symptoms of current low-level panic.

Other risk factors have been identified. Experiencing physical or sexual abuse during childhood, for example, leads to a significantly increased risk of panic attacks in adulthood. Genetic factors play a part too, with heritability estimated at around 40%. In fact, it’s been suggested that panic disorder may be the most heritable of all the anxiety disorders. Like those other disorders, however, we’re nowhere near knowing exactly which genes help cause panic disorder.

Although the debate between proponents of the biological and psychological perspectives on panic continues, the weight of evidence favours the latter. Perhaps the most telling piece of this evidence is the remarkable success of psychologically based treatments for panic disorder, especially cognitive behaviour therapy. These treatments focus on changing the way people think about their physical sensations, their panics, and their ability to cope with them. If panic were primarily a biological reflex, altering patterns of thought would presumably make less therapeutic difference.