Next, Clark and his colleagues asked if activating this habit with words would induce panic. All the subjects read a series of word pairs a aloud. When panic patients got to "breathless-suffocation" and "palpitations-dying," 75 percent suffered a full-blown panic attack right there in the laboratory. No normal people had panic attacks, no recovered panic patients (I'll tell you more in a moment about how they got better) had attacks, and only 17 percent of other anxious patients had attacks.

The final thing Clark told us was the "breakthrough" that Rachman had promised.

"We have developed and tested a rather novel therapy for panic," Clark continued in his understated, disarming way. He explained that if catastrophic misinterpretations of bodily sensation are the cause of a panic attack, then changing the tendency to misinterpret should cure the disorder. His new therapy was straightforward and brief:

Patients are told that panic results when they mistake normal symptoms of mounting anxiety for symptoms of heart attack, going crazy, or dying. Anxiety itself, they, are informed, produces shortness of breath, chest pain, and sweating. Once they misinterpret these normal bodily sensations as an imminent heart attack, their symptoms become even more pronounced because the misinterpretation changes their anxiety into terror. A vicious circle culminates in a full-blown panic attack.

Patients are taught to reinterpret the symptoms realistically as mere anxiety symptoms. Then they are given practice right in the office, breathing rapidly into a paper bag. This causes a buildup of carbon dioxide and shortness of breath, mimicking the sensations that provoke a panic attack. The therapist points out that the symptoms the patient is experiencing--shortness of breath and heart racing--are harmless, simply the result of overbreathing, not a sign of a heart attack. The patient learns to interpret the symptoms correctly.

"This simple therapy appears to be a cure," Clark told us. "Ninety to 100 percent of the patients are panic free at the end of therapy. One year later, only one person had had another panic attack."

This, indeed, was a breakthrough: a simple, brief psychotherapy with no side effects showing a 90-percent cure rate of a disorder that a decade ago was thought to be incurable. In a controlled study of 64 patients comparing cognitive therapy to drugs to relaxation to no treatment, Clark and his colleagues found that cognitive therapy is markedly better than drugs or relaxation, both of which are better than nothing. Such a high cure rate is unprecedented.

How does cognitive therapy for panic compare with drugs? It is more effective and less dangerous. Both the antidepressants and Xanax produce marked reduction in panic in most patients, but drugs must be taken forever; once the drug is stopped, panic rebounds to where it was before therapy began for perhaps half the patients. The drugs also sometimes have severe side effects, including drowsiness, lethargy, pregnancy complications, and addictions.

After this bombshell, my own "discussion" was an anticlimax. I did make one point that Clark took to heart. "Creating a cognitive therapy that works, even one that works as well as this apparently does, is not enough to show that the cause of panic is cognitive:" I was niggling. "The biological theory doesn't deny that some other therapy might work well on panic. It merely claims that panic is caused at the bottom by some biochemical problem."

Two years later, Clark carried out a crucial experiment that tested the biological theory against the cognitive theory. He gave the usual lactate infusion to 10 panic patients, and nine of them panicked. He did the same thing with another 10 patients, but added special instructions to allay the misinterpretation of the sensations. He simply told them: "Lactate is a natural bodily substance that produces sensations similar to exercise or alcohol. It is normal to experience intense sensations during infusion, but these do not indicate an adverse reaction." Only three out of the 10 panicked. This confirmed the theory crucially.

The therapy works very well, as it did for Celia, whose story has a happy ending. She first tried Xanax, which reduced the intensity and the frequency of her panic attacks. But she was too drowsy to work, and she was still having about one attack every six weeks. She was then referred to Audrey, a cognitive therapist who explained that Celia was misinterpreting her heart racing and shortness of breath as symptoms of a heart attack, that they were actually just symptoms of mounting anxiety, nothing more harmful. Audrey taught Celia progressive relaxation, and then she demonstrated the harmlessness of Celia's symptoms of overbreathing. Celia then relaxed in the presence of the symptoms and found that they gradually subsided. After several more practice sessions, therapy terminated. Celia has gone two years without another panic attack.

EVERYDAY ANXIETY

Attend to your tongue--right now. What is it doing? Mine is swishing around near my lower right molars. It has just found a minute fragment of last night's popcorn (debris from Terminator 2). Like a dog at a bone, it is worrying the firmly wedged flake.

Attend to your hand--right now. What's it up to? My left hand is boring in on an itch it discovered under my earlobe.

Your tongue and your hands have, for the most part, a life of their own. You can bring them under voluntary control by consciously calling them out of their "default" mode to carry out your commands: "Pick up the phone" or "Stop picking that pimple." But most of the time they are on their own. They are seeking out small imperfections. They scan your entire mouth and skin surface, probing for anything going wrong. They are marvelous, nonstop grooming devices. They, not the more fashionable immune system, are your first line of defense against invaders.