Article: Miracle Drug, Poison or Placebo?

By Maia Szalavitz for MSN Health & Fitness
MSN Health & Fitness Exclusive

Modern antidepressants have been blamed for deadly shooting rampages and violent suicides. At the same time, they’ve been hailed as miracle drugs that transform baleful Eeyore-types into bouncing Tiggers. Now the latest review of the research claims that the effects of the drugs are only marginally different from those of placebos or sugar pills. It seems impossible that the same substances in the same dosage ranges could simultaneously be poison, miracle drug and placebo. But the diversity of responses is remarkable—and it points to the possibilities and pitfalls of personalized medicine. For example, Stacy*, a 48-year-old woman who works in public relations in Ohio, describes her experience with Zoloft like this: “It felt like water after being in the desert. It wasn’t an experience of elation or anything bi-polar … I’m far happier, more confident, far more relaxed.” Lisa*, a 33-year-old business consultant from Maryland, had experienced severe suicidal thoughts as early as kindergarten. She says of taking Effexor, “My entire life is different and I finally feel like a normal person with normal emotions. These days I can honestly say I am a happy, well adjusted person.” But JoAnne*, a 35 year-old educator and dancer living in the Washington, D.C. area, reported that both Zoloft and Prozac produced muscle weakness and excessive sweating—and no benefits. And Bernice*, a 53-year-old science journalist in California, described her experience with a Prozac-like antidepressant this way: “It made me feel disconnected from myself and my family, so that I no longer felt any empathy and did not really care what happened to them or to me. It was a terrifying sensation of flatness and I definitely felt depressed and hostile in a way that I had never felt before.” Bizarre experiences abound as well: Bernice had “a vivid nightmare of being shot in the head,” and the sensation she felt of dripping blood did not immediately disappear on awakening. Others report elimination of sexual desire, weight loss, weight gain, heart palpitations and elevated blood pressure. *not their real names Same Drug, Different Results How can similar—or even identical—drugs have such different effects? The explanation sheds light on some of the key questions that complicate research on depression and emphasizes the need for a science of personalized medicine. “These are average changes,” explains Richard Tranter, M.D., consultant psychiatrist with the North West Wales NHS Trust in Great Britain. “They may actually mask wide variations in how individuals are responding.” “If you look at a large scale trial,” Tranter continues, “You may get a slight shift overall. That could mean on the whole, everyone is a little bit better compared to placebo but it could also mean that within the group, some did really, really well and some did OK and some got worse, but overall, it looks like a small shift in the right direction.” Two other complications are noted by psychiatrist Peter Kramer, author of Listening to Prozac. First, early data—of the kind included in the meta-analysis which found antidepressants no better than placebo—often fails to show results for drugs that later prove very effective. “We have medications that show effectiveness in post-stroke depression, in the elderly with multiple brain lesions, in people with chronic, intractable depression. Why do they only not work in the early trials?” he asks. Second, people get excellent care in clinical trials, which makes the placebo effect especially strong. This means it’s harder to show a difference, even if a drug works well. “We may be losing effective substances, not approving ineffective ones,” Kramer notes. A Complex Challenge: Finding the Right Drug So, how can you find an antidepressant that will work, not make you feel worse? That’s the billion dollar question that pharmaceutical companies, psychiatrists and researchers are racing to answer. The complexities, however, are daunting. A large proportion of human genes code for brain development and function, and they interact in complex ways not only with each other, but with environmental influences like stress, parenting and social support. One intriguing study by Tranter and colleagues involved giving antidepressants to healthy people. The study compared the results of a drug similar to Prozac, which affects the serotonin system in the brain, to the effects of another drug, which affects the norepinephrine system. Serotonin—a neurotransmitter that sends messages from one brain cell to another—is involved with regulating mood. Norepinephrine (also called noradrenaline) helps mediate attention and the brain’s response to stress. The antidepressants in the study affect these transmitters by keeping them activated for longer than usual. “One of the effects of [the Prozac-like drug] was what you might call a ‘serenic’ effect,” Tranter says. “It’s this feeling of being aware of stressful things going on, but they don’t bother you much.” Some people liked this effect—as did the people around them, who found them easier to get along with. Others, however, reported this same experience, but found it unpleasant. “They felt disconnected from their emotions, unpleasantly numb, and they didn’t like it at all,” Tranter says. The participants in Tranter’s study had been given personality tests beforehand. Some of the traits measured on these tests are strongly influenced by genetics, reflecting inborn temperamental differences. People who liked one of the drugs tended to dislike the other—and those high in a trait called “harm avoidance,” (basically, a sense of caution and worry) typically preferred the norepinephrine-affecting drug and disliked the serotonin drug. Other studies have found possible links between particular genes and particular responses: A gene associated with men’s ability to control anger was found to triple the risk that men would develop suicidal thoughts if given the drug citalopram (Celexa). The long version of a serotonin gene associated with the risk of developing depression following early life trauma was associated with a better or faster response to Prozac-like drugs, compared to the shorter version of that gene.Another serotonin-related gene was found to affect whether or not people experienced negative sexual side effects from one of these medications. Variations in genes for one of the brain’s opioid receptors have also been associated with positive responses to citalopram (Celexa). Other Factors Unfortunately, most of these results failed to be replicated when researchers looked for them again. “People find predictors of response all the time but they are almost never replicated,” says Roy Perlis, M.D., an assistant professor of psychiatry at Harvard Medical School. Genetic diversity may complicate the story. For example, one gene associated with positive response to antidepressants in whites was associated with lack of response in Japanese and Korean people. The environment matters too—one study found that people who had been abused in childhood were less likely to respond to medicationsand more likely to be helped by talk therapy. To complicate matters even further, it’s not just brain genes that affect drug response, but also a system of enzymes in the liver in which there is also great genetic variation. For example, the enzyme CYP450 2D6 metabolizes Prozac and some related drugs and there are numerous variants of this one enzyme alone that can affect the way these drugs work. People whose 2D6 breaks down Prozac too fast may get no effect from ordinary dosages of the drug. Conversely, those whose enzyme works slowly may have severe side effects even at low dosages. A genetic test that is now commercially available can determine which variant someone has— but there’s another wrinkle. Dietary factors significantly affect the actions of various enzymes. You may have heard of grapefruit juice negatively affecting the impact of drugs, but char-broiled meat, broccoli, star fruit, alcohol and tobacco also affect the response to medications. In addition, interactions between numerous medications can make things even less predictable. “For the majority of people, these tests are not useful for antidepressant prescribing,” Perlis concludes. However, for people with bipolar disorder or psychotic depression who use antipsychotic medications, the tests can be important, particularly in terms of preventing serious side effects. “It’s a nice idea,” concludes Tranter. “A lot of people are very interested in using genetics to predict the response to medication, but it’s too early to know what it will yield.” Treating Depression Step By Step But although trial and error and doctors’ clinical intuition are still all we really have to go on in antidepressant choice, a large study examining treatment of depression in real-world practice found that most people did benefit significantly by the end of the trial. Unlike a clinical trial, which usually just compares a drug to placebo or another drug, this trial involved trying different drugs sequentially if the first ones didn’t work for particular patients. “Two-thirds got better by the end of the steps,” says Madhukar Trivedi, M.D., professor of psychiatry at the University of Texas Southwestern Medical School. “And by better, I mean that almost all the symptoms were gone—they were in remission. Not just showing improvement.” Sound Off: Share Your Experience With Antidepressants More on MSN Health & Fitness: • Do Antidepressants Lose Their Power? • Sexual Side Effects of Antidepressants • Why Do Antidepressants Make Me Anxious? • The Truth About Painkillers • 10 Deadliest Drugs Popular Slide Shows on MSN Health & Fitness: • Unusual Symptoms, Common Ailments • What Your Hair Says About Your Health • A Day at the Dental Spa Maia Szalavitz is a freelance journalist and senior fellow at media watchdog, is co-author with Bruce D. Perry, MD, PhD, of The Boy Who Was Raised as a Dog and Other Stories from a Child Psychiatrist’s Notebook: What Traumatized Children Can Teach Us About Loss, Love and Healing (Basic, 2007). URL:>1=31009