REPOSTED (orig. March 1, 2008) —
Five days after the February 2008 shooting rampage at Northern Illinois University, the Chicago Tribune ran a headline that caught my eye: “Doctors: Prozac, Violence Rarely Linked.” Given the ambiguous grammar, it was hard to tell if the headline was warning doctors about links between the antidepressant and violence or quoting doctors as saying we needn’t be concerned about such links (“rarely linked”). A complete ruling on the matter would of course be reassuring. It might calm nationwide jitters that the cocktail of drugs—Prozac, Xanax, and Ambien—that Steven Kazmierczak had abruptly stopped taking, days before he went on his shooting spree, was connected with the violence that followed.
The Tribune headline implied that although there was a connection between Prozac and violence, it was not a significant one. When I read the article, however, I learned that Jeremy Manier, its author, wasn’t so sure. Quoting deep-seated concern by experts at renowned local hospitals, including the University of Chicago Medical Center and Northwestern’s Feinberg School of Medicine, he wrote: “About one-fifth of people who halt a course of Prozac-like drugs report symptoms associated with a condition known as discontinuation syndrome, which can include abdominal pain, dizziness, crying spells, irritability and even a sensation similar to an electrical shock in the patient’s arms or legs.”
A disturbingly large number of studies corroborate Manier’s statement. Warnings from experts about a host of problems tied to ending S.S.R.I. (selective serotonin reuptake inhibitor) treatment have spotlit other areas of concern about this class of medication—areas that will need exhaustive investigation before they can be considered resolved. These include the drugs’ effectiveness relative to placebo, and their published track record. The New England Journal of Medicine recently disclosed that the drugs’ successes have been consistently exaggerated over a period of seventeen years. As a result of such distortion, drugs like Zoloft appear in the pharmacological literature to be 70 percent more effective than the data tell us they actually are.
Given the prevalence of discontinuation syndrome and the erratic effect of mixing different classes of medication, no one should rush to judgment. Many variables are in play, including how differently people respond to S.S.R.I. medication; how severely disordered they were before taking it; and whether they are using it in combination with other drugs like Xanax and Ambien. But the sheer amount of guesswork surrounding such combinations, and how common they have become, should give us pause. The current concern about S.S.R.I. medication intensifies when one considers how many patients are cycling through other kinds of drugs at the same time, interactions that are not fully known or studied.
Not all discontinuation symptoms result in self-inflicted or externalized violence, a point I mention rather than minimize. But the reason the Food and Drug Administration added black-box warnings to S.S.R.I.s, alerting physicians to the risks of prescribing them to children and adolescents, was concern about their spotty track record and, in particular, indications of a link to violence. Numerous studies over the years pointed to a significant number of patients on the drugs who either attempted suicide or obsessed about doing so. The agency decided that it needed to take action. No one wanted a disturbing pattern to balloon into an established trend.
The drug companies want to relabel these symptoms as a resurgence of the original disorder. The problem they face in doing so is that discontinuation syndrome is entirely drug related. Prozac’s maker, Eli Lilly, has fought several protracted legal battles trying to dislodge evidence that its psychotropic is linked to violence—and Lilly is not the only drug maker that has had trouble making its case convincing.
As the Tribune’s Manier reminds us, Eric Harris and Dylan Klebold, the shooters at Columbine High School, abruptly stopped taking the same class of antidepressant medication days before they opened fire on their classmates. Jeff Weise, the Red Lake High School killer in Minnesota, was taking Prozac before he killed nine people and then himself. Pekka-Eric Auvinen, the eighteen-year-old who began shooting in Jokela High School, Finland, had a history of S.S.R.I. use. According to investigators, so did Seung-Hui Cho, who killed thirty-two people at Virginia Tech and wounded dozens more.
The list of other killings involving S.S.R.I. psychotropic medication is distressingly long. It includes Michael McDermott, the software engineer who went on a rampage in Massachusetts, killing nine; Byran Uyesugi, who shot seven of his colleagues in Hawaii; and Charles Carl Roberts IV, who assassinated five Amish school girls before shooting himself.
Such incidents may amount to nothing more than an awful set of coincidences. But many people are sufficiently alarmed by signs of a pattern to suggest that the repeated use of psychotropic medication is involved—that drugs are part of the problem here, rather than, as commonly assumed, its solution. With the pattern that is emerging, the standard defense by psychiatrists and drug companies—that patients’ quitting medication simply demonstrates how much it was needed in the first place—holds less water, especially in light of the black-box warnings, added to these drugs by the F.D.A, that indicate they can increase suicide ideation in patients, including in those who stop taking their medication abruptly.
Was the Tribune headline correct, then, when it called Prozac and violence “rarely linked”? The answer to that question depends in large measure on how one defines “rarely” and “linked.” Some would say that “rarely” is not a word to generate much concern, because the number it refers to is statistically insignificant. Yet according to the International Review of Psychiatry, more than 67.5 million Americans—almost a fifth of the country—have taken a course of S.S.R.I. medication. Twenty percent of them constitutes a sizable crowd—roughly the metropolitan populations of New York City and Los Angeles combined.
What about “linked”? Interestingly, Manier’s statement about the one in five patients who experience discontinuation syndrome on Prozac corroborates the words of Paul N. Jenner, who in 1998 distributed a confidential memo on this subject to executives at GlaxoSmithKline. Jenner was at the time the company’s Director and Vice President of Worldwide Strategic Product Development, so he was well placed to warn that Paxil too presented a “20 percent relapse rate.” Nevertheless, when highlighting what his report dubbed “Issues Management,” Jenner assured colleagues that “our highly skilled sales and marketing efforts” would spin “the discontinuation issue,” deflecting negative publicity by playing up Paxil’s “flexibility and control.”
In the report, at least, Jenner voiced not a shred of concern that one in five patients on Paxil was experiencing mild-to-serious side effects. (A later health report from GSK would list these as including risk of coma, birth defects, blood aggregation problems, and renal failure.) He was far more worried about the “fight for market share,” with competitors like “Lilly and Pfizer resorting to aggressive tactics to undermine Seroxat/Paxil’s growth.”
“Lilly,” Jenner complained, was “currently focusing on the issue of discontinuation, on trying to turn a disadvantage into an advantage by playing to the supposed strength of [Prozac’s] long half-life . . . providing an in-built tapering mechanism. This is clearly a marketing ploy,” he concluded, “already seen through by most psychiatrists, and a sign of desperation in the fight for market share.”
Coming from the makers of Paxil, such complaints might sound like the pot calling the kettle black. But Jenner was right that Lilly had tried to sugarcoat grave concern among clinicians and researchers about Prozac’s discontinuation syndrome—concern that helped prompt the FDA to take action, but that has not gone away, because discontinuation syndrome afflicts all ages.
As there are still so many “unknown unknowns” about S.S.R.I. antidepressants, what is needed now is a frank, open dialogue about the evidence we do have, including the efficacy and erratic effects of these medications when combined with other drug treatments. Longer clinical trials representing the full spectrum of patient reaction over six months—rather than, as is common, two weeks—would give us a clearer picture of how the brain and central nervous system react when patients come off this class of medication. As the Tribune article and a litany of studies make disturbingly clear, the evidence is mounting that these psychotropics have far more worrisome, unpredictable effects than large numbers of prescribers and drug makers would have us believe.
Christopher Lane, Professor of English at Northwestern University, is the author most recently of “Shyness: How Normal Behavior Became a Sickness.”