When I was researching Anatomy of an Epidemic, a book that Crown will publish on April 13 [already published; this piece was written in 2011], I was curious, in particular, to see how psychiatry rebuilt the market for benzodiazepines in the 1980s. During the 1970s, both the U.S. and British governments concluded that benzodiazepines caused physical dependence and thus capable of causing a great deal of harm, and for a time, the prescribing of these drugs in the U.S. did noticeably decline, from 103 million prescriptions in 1975 to 71 million in 1980. But then the prescribing numbers began to stabilize and eventually increase, such that 83 million prescriptions for benzodiazepines were written in 2007, which isn’t all that far from the number written at the height of the Valium craze in the early 1970s.
Here’s how the market for benzodiazepines was revived.
In 1981, Upjohn brought a potent new benzodiazepine to market, Xanax, and this helped stabilized use of the drugs for “anxiety.” Next, Upjohn conducted a study of Xanax as a treatment for panic disorder, which the American Psychiatric Association had newly identified as a discrete condition. Then, in May 1988, Upjohn investigators announced that the study had shown Xanax to be a safe and effective treatment for this newly identified “illness.”
At least at first glance, it seemed that science had indeed shown Xanax to be helpful. The FDA approved it as a treatment for panic disorder, and the American Psychiatric Association touted its benefits to the public. At the same time, NIMH identified panic disorder as a priority concern and in 1991 sponsored a conference on it, with its panel of experts designating “high potency benzodiazepines”—this would be Xanax—as one of the two “treatments of choice.”
Yet, a close look at the 14-week study conducted by Upjohn shows that this drug, rather than provide a benefit to patients, caused a significant amount of harm.
Benzodiazepines are known to work quickly, and at the end of four weeks, 82 percent of the Xanax-treated patients were “moderately improved” or “better,” versus 43 percent of the placebo group. However, during the next four weeks, the placebo patients continued to improve while the Xanax patients did not, and by the end of the eighth week, there “was no significant difference between the groups” on most of the rating scales, at least among the patients still in the study. The Xanax patients also experienced a variety of troubling side effects: sedation, fatigue, slurred speech, amnesia, and poor coordination.
At the end of eight weeks, the patients were tapered from their medication (Xanax or placebo) for four weeks and then followed while medication-free for another two weeks. Thirty-nine percent of those withdrawn from Xanax “deteriorated significantly,” their panic and anxiety skyrocketing to such an extent they had to start taking the drug again. Thirty-five percent suffered “rebound” panic and anxiety symptoms more severe than when the study began, and an equal percentage suffered a host of debilitating new symptoms, including confusion, heightened sensory perceptions, depression, a feeling that insects were crawling over them, muscle cramps, blurred vision, diarrhea, decreased appetite, and weight loss.
In sum, at the end of the 14-week study, the drug-exposed patients were much worse off than the placebo group. They were more anxious, more panic stricken, and doing worse on a “global scale” that assesses overall well-being. Forty-four percent had been unable to get off the drug, on their way to a lifetime of physical dependence.
So how Upjohn’s investigators turn this study into evidence that Xanax was a “safe and effective” treatment for panic disorders? In their published reports, the Upjohn investigators focused on the four-week results, when the drug-treated patients were doing better than the placebo group. They acknowledged that many Xanax-treated patients had fared poorly when the drug was withdrawn, but they reasoned that this showed that the drug had been used for too short a period, and the withdrawal done too abruptly. “We recommend that patients with panic disorder be treated for a longer period, at least six months,” they said.
Such is the story of how the market for benzodiazepines was revived in the 1980s. Upjohn conducted a trial that told of drug-treated patients ending up much worse than the placebo patients by the end of the 14 weeks, but its investigators instead reported that the trial had shown Xanax to be safe and effective, a conclusion they supported by focusing on the four-week results. The APA and the NIMH echoed this conclusion, and so sales of Xanax took off, with this drug, in 1992, becoming the fifth most frequently prescribed medication in the United States.
The podcast below (starting at the 16:14 mark) features an interview with Robert Whitaker where he expands in greater detail on this topic.