Content provided by the Faculty of the Harvard Medical School

In brief: The danger of playing it safe

In brief

The danger of playing it safe

People with anxiety disorders — social phobia, panic disorder, generalized anxiety — find ways to avoid the dangers they imagine they foresee. They may repeatedly check their pulse, constantly ask for reassurance, or carry medication and a cell phone wherever they go. They often ward off anxiety this way, to their ultimate disadvantage. An experiment conducted at the University of Texas shows that having this kind of protection available prevents recovery, whether the protection is actually used or not.

Seventy-two undergraduates with claustrophobia were assigned at random to one of five groups, three of which received exposure treatment. They were told that claustrophobia results from fear of being trapped or suffocated and that the fear is perpetuated by avoidance. Then they repeatedly performed an exercise that required them to stay for 5 minutes at a time in a small chamber — 6 feet high, 2 feet wide, and less than 2 feet deep — with black walls, no lights, and a closed and covered window.

The first group was given no way out — standard exposure therapy. In the second group, subjects were told that they were expected to open the window or the door or talk to an experimenter by radio at some point. In the third group, they were told to use these safety strategies only if they must.

The fourth group received a placebo. They wore goggles and a headset that emitted sounds and lights in a pattern that is supposed to have a relaxing effect. The device is not regarded as a treatment for phobias. The fifth group was simply put on a waiting list.

Improvement was judged by the degree of fear subjects reported when they entered the black chamber again after treatment. In the no-safety group (exposure therapy only), 94% improved significantly. In the groups provided with loopholes — whether they were told that using the loophole was expected or that they should use it only if they had to — the improvement rate was about 45%. In the placebo control group it was 25%, and there was no change in the students put on a waiting list. Relapse rates after two weeks were low and statistically similar in all groups.

All of the subjects in the second group used one of the escape methods or sought reassurance from outside. None of those in the third group did, but it made no difference. Simply knowing that there was a way out made recovery less likely, because they were not really confronting the threat. To prevent patients with phobias from playing it safe in everyday life, the authors say, therapists should explain this clearly and tell them specifically which safety tricks and techniques they have to avoid.

Powers MB, et al. “Disentangling the Effects of Safety- Behavior Utilization and Safety-Behavior Availability During Exposure-Based Treatment: A Placebo-Controlled Trial,” Journal of Consulting and Clinical Psychology (May–June 2004): Vol. 72, No. 3, pp. 448–54.

E-mail This | Print This

Last updated:	August 21, 2006