06.14.11

Do Antidepressants Make You Sad?

Some new research suggests people who take antidepressants are more likely to experience relapses of depression.

In the pendulum-swinging debate over the effectiveness of antidepressants, the latest findings by some researchers will not be good news for users and supporters of the ubiquitous meds: That used over the long term, antidepressants may "do more harm than good," and once discontinued are more likely to cause a relapse in the patient—more so than with a patient taking a placebo.

"There are data now that suggest that the longer the period of administration, the more likely the patient will relapse," says Dr. Giovanni Fava, a professor of clinical psychology at the University of Bologna in Italy, who has been looking at this possibility for nearly 20 years. His latest article, currently in press, is to be published in the journal "Progress in Neuro-Psychopharmacology and Biological Psychiatry."

And Dr. Irving Kirsch, a professor of psychology at the University of Hull, in the U.K., and the author of The Emperor's New Clothes: Exploding the Antidepressant Myth who spent 15 years examining sheaths of data—much of it unpublished—from clinical trials of antidepressants, has concluded that antidepressants are, in essence, hardly better at alleviating depression than placebos; in fact, he argues that the widely accepted belief that depression is, at bottom, a chemical imbalance, is a misconception. 

Kirsch is in the midst of looking at data that align with Fava's argument: taking antidepressants—especially SSRIs—might increase the risk of relapse. Kirsch emphasizes that what he has found so far is not conclusive. But he says: "I've noticed two things. One is, if you look at clinical trials that follow patients for up to a year after they've been successfully treated in the placebo arm and you keep treating them with placebo and you see what happens, you get about a 20 percent relapse in that group. Most of them stay better, actually, so that's encouraging. If you look at trials who've gotten better on drug and then switched to placebo to then see how many relapse—they've been in a trail for four to eight weeks on antidepressants and are then switched to placebo—about 50 percent relapse."

Questions about how well antidepressants work in the treatment of depression have been in the mix for several years; large-scale studies, like the influential STAR*D trials, have turned up lackluster results, showing antidepressants to be little more effective than placebos in combating mild to moderate cases of depression. And a recent spate of books questioned how and why we've gotten so entangled in a pharmaceutical existence.

Antidepressant drugs, for me, they're like antibiotics, If they're used properly, they're very good. But if they're used when it's not necessary, or for the wrong indication, or if they're prolonged for too long, then they're not helpful.

Fava's latest article, currently in press, is to be published in the journal Progress in Neuro-Psychopharmacology and Biological Psychiatry. His research is not based on his own experimental trials, but rather on an examination of the existing depression literature. Looking at a swath of studies on the effects of antidepressants, Fava found several troubling patterns, including an indication that they not only fail to protect against another episode of depression but, "on the contrary," could result in relapse.

For instance, looking at one study (from Gardarsdottir et al, 2009), Fava reports that in a sample of more than 9,000 patients treated with SSRI's, those who took medication for a year or longer were up to 23 percent more likely to have a second episode of major depression than those who stopped taking the pills within the first six months.

Fava also notes that the longer you take antidepressants, the less effective they become. "We have data showing that there is loss of clinical effect in at least 1/3 of patients after only a few months of therapy. So the drugs do not really work as well as they did at the beginning," he says. "And what happens is at that point, you may add other drugs or you may increase dosage—and for a certain time, this works, but then you generally get back to the way things were."

Fava makes a telling comparison. "Antidepressant drugs, for me, they're like antibiotics ," he says. "If they're used properly, they're very good. But if they're used when it's not necessary, or for the wrong indication, or if they're prolonged for too long, then they're not helpful." Further, he observes, discontinuing antidepressants is often accompanied by withdrawal symptoms that, in some cases, do not go away.

While the data surrounding relapse is, for Fava, of central importance, for Madhukar Trivedi, the director of the mood disorder research program at University of Texas South Western medical school, says the evidence is not yet there to make the case. One possible confounding problem, Trivedi argues, is the severity of the depression itself. Are patients who are staying on the drugs long-term also the patients whose depression is so severe that they'd be likely to relapse anyway? There've not yet been enough long-term experimental trials to know, he argues.

"I think that it's a question worth having, but I don't think we have a definitive answer that means yes, the exposure to the antidepressant five years back or 10 years back caused it," says Trivedi, who adds that, for him, "this relapse problem is kind of academically and intellectually interesting, but in general I think that the patient's suffering really comes from not having good enough outcome from the current treatments they're on. And that is not a relapse problem, that is a current disease problem." 

Fava posits a hypothesis that could explain his collection of findings—he calls it "the oppositional model of tolerance." What it means, in essence, is that taking antidepressants causes the brain and body to compensate in the opposite direction and induce the very problems they've been enlisted to solve. The exact mechanisms in the brain on which antidepressants might work to make relapse more likely are not, by any means, clear. 

Kirsch's explanation is similar to Fava's "oppositional model of tolerance." 

"This can make some sense in terms of the hypothesis that giving people serotonin can produce a kind of compensatory response in the body," Kirsch says, "such that—and I don't know how—but it may be producing a change in the brain that then leaves the person more vulnerable to becoming depressed, one way or another."

If it's true, as Fava hypothesizes, that antidepressants predispose patients to a recurrence of the disorder, it is also true that depression itself—untreated—predicts at least a second major depressive episode, if not multiple episodes. Here, Fava is emphatic in his message that antidepressants are important—"life-saving"—in the short-term treatment of severe cases of depression.

The state of depression research in a country some refer to as the "Prozac nation" is rife with disagreement and uncertainty; this latest strand of doubt over the issue of relapse exemplifies the problem.

In the meantime, Fava asks, why is it that so many psychiatrists are incredulous that successfully treating depression should involve not only the right pill but also the appropriate change in thinking? And a shift in how we proceed through our daily lives? When it comes to cardiovascular health, cholesterol levels, or diabetes, doctors emphasize lifestyle changes as well as prescription pills, Fava points out. Yet, he says, "psychiatrists are pretty alone in the medical literature having faith only in drugs, and not in changing the attitude of the patients."