For decades, the public and most mental health professionals have felt that antidepressant medications are a magic bullet for depression. Beginning in the late 1950s, antidepressants ushered in an era of safe, reliable and reasonably affordable treatment that often produced better results than the psychotherapies of the day. As the compounds rose in popularity, many physicians came to view psychotherapy alone as ineffective and as little more than a minor adjunct when combined with medication.

This is no longer the case, if it was ever true. Contrary to prevailing wisdom, recent research suggests that several focused forms of psychotherapy may be as effective as medication, even when treating more severe depressions. Moreover, the newer psychotherapies may provide advantages beyond what antidepressants alone can achieve. Nevertheless, pharmaceutical therapy remains the current standard of treatment, and effective new options are being added all the time.

These trends are important to examine because depression exacts a significant toll on society as well as individuals. Depression is one of the most common psychiatric disorders and is a leading cause of disability worldwide. The impact of mood disorders on quality of life and economic productivity matches that of heart disease. Depression also accounts for at least half of all suicides. The efficacy of antidepressants has been established in thousands of placebo-controlled trials. The newer ones are safer and have fewer noxious side effects than earlier compounds. About 50 percent of all patients will respond to any given medication, and many of those who do not will be helped by another agent or a combination of them. Not everyone responds, however, and many who do would prefer not to have to take the pills. Quietly over the years, newer psychotherapeutic techniques have been introduced that may be just as good at alleviating acute distress in all but the most severely depressed patients. And some of the therapies provide advantages over medication alone, such as improving the quality of relationships or reducing the risk that symptoms will return after treatment is over.

This last revelation is significant because many people who recover from depression are prone to succumb again. The illness is often chronic, comparable to diabetes or hypertension, and patients treated with medication alone may have to remain on it for years, if not for life, to prevent symptoms from returning. Moreover, combining treatments--prescriptions to reduce acute symptoms quickly and psychotherapy to broaden their effects and to prevent symptoms from returning after treatment terminates--may offer the best chance for a full recovery without recurring problems.

Remission or Relapse
Our conclusions refer mainly to the condition termed unipolar disorder. Depression comes in two basic forms: The unipolar type involves the occurrence of negative moods or loss of interest in daily activities. In the bipolar form, commonly known as manic-depression, patients also experience manic states that may involve euphoria, sleeplessness, grandiosity or recklessness that can lead to everything from buying sprees to impulsive sexual adventures that later bring regret.

Bipolar disorder shows up in only 1 to 2 percent of the population and is usually treated with mood-stabilizing medication such as lithium. In contrast, about 20 percent of women and 10 percent of men suffer from unipolar depression at some time in their lives.

The treatment of unipolar depression typically progresses through three phases, determined by changes in the patient's intensity of symptoms. These are usually measured by clinical ratings such as the Hamilton Rating Scale for Depression. Seriously depressed patients in the acute phase often report feeling down much of the time. They have lost interest in formerly pleasurable activities, and they may have difficulty sleeping, changed appetite, and diminished libido. They may feel fatigued or worthless, and they may entertain recurrent thoughts of death or suicide. The goal of treatment is to relieve symptoms. "Remission" is reached when someone is fully well.

Even when in remission, however, patients may still have an elevated risk for the return of symptoms. It is common practice to encourage patients to stay on medication for at least six months following the initial remission. The return of symptoms soon after remission is called a relapse. In this sense, treating depression with drugs may be like treating an infection with antibiotics; a patient must take the medication beyond the point of first feeling better, to fully prevent the original problem from coming back. This effort to forestall relapse is called continuation treatment and typically lasts at least six to nine months beyond the point of remission.

Those who pass the point at which the treated episode is likely to return are said to have recovered. But even then, they might experience a new episode; people with a history of depression are three to five times more likely to have an episode than those with no such history. A new episode is considered a recurrence. To protect against recurrence, many patients are kept in ongoing maintenance treatment, typically medication but sometimes with psychotherapy. But once patients are off medication, having been on it does nothing to reduce subsequent risk for recurrence. Therefore, patients with a history of multiple episodes are usually advised to stay on medication indefinitely.

Although the scope of depression can vary widely, there are only a few prevailing treatments. Most of the leading antidepressants fall into three main classes: monoamine oxidase inhibitors (MAOIs), tricyclic antidepressants (TCAs), and selective serotonin reuptake inhibitors (SSRIs), such as Prozac and Paxil. Each class has a slightly different action and different side effects and is prescribed based on a patient's history, the likelihood of certain complications, and cost. Although about equally effective in a general population, some medications are more efficacious than others for specific types of depression. In general, the older MAOIs and TCAs carry greater risk of side effects than the SSRIs. But the SSRIs do not always work, especially for more severely depressed patients, and they are more expensive.

Despite the widespread use of antidepressants, their actions are not fully understood. They work in part by affecting the neurotransmitters (signaling molecules in the brain) norepinephrine, serotonin and dopamine, which are involved in regulating mood, primarily by blocking the reuptake of these neurotransmitters into the neurons that secrete them. Yet this action cannot fully explain the effects, and it is quite likely that the compounds drive a subsequent cascade of biochemical events. Many people who do not respond to one antidepressant will respond to another or to a combination.

New psychotherapy methods have proved as effective as medication, although they are still not as extensively tested. The programs include interpersonal psychotherapy (IPT), which focuses on problems in relationships and helps patients lift the self-blame common in depression. Developed in the 1970s, IPT has performed well in trials but has only begun to enter clinical practice. Studies do show, however, that when IPT is paired with medication, patients receive the best of both worlds: the quick results of pharmaceutical intervention and greater breadth in improving the quality of their interpersonal lives.

Cognitive and behavioral therapies, collectively known as CBT, also compare well with medication in all but the most severely depressed patients--and they can benefit even those people if they are administered by experienced therapists. Most exciting is that CBT appears to have an enduring effect that reduces risk of relapse and perhaps recurrence. Even the most effective of the other treatments rarely have this type of long-lasting benefit. Cognitive therapy is perhaps the most well established CBT approach. It teaches patients to examine the validity of their dysfunctional depressive beliefs and to alter how they process information about themselves. Behavioral therapy had lost favor to the cognitive approaches, but it, too, has done well in recent trials and is undergoing a revival.

Which Way to Turn
It is not possible to simply say whether medication or psychotherapy is "better" for depressed patients. But many studies have reached interesting conclusions about the approaches when they are applied across the illness's three phases: the acute symptoms at onset, the months of continuation treatment to forestall relapse, and the maintenance of health for years to come.

Among patients who take antidepressants during treatment for acute symptoms, about half show a 50 percent drop in symptom scores on rating tests over the first four to eight weeks. About one third of those patients become fully well (remission). Not all the improvement can be attributed to pharmacology, however. In pill-placebo control experiments, placebos can achieve up to 80 percent of the success rate of active medication, probably by instilling in patients hope and the expectation for change. The placebo effect does tend to be less stable over time and smaller in magnitude in more severe or chronic depressions. A major problem with acute-phase therapy, however, is that many stop taking their medication--primarily because of side effects--before therapists can clearly tell if the agents are working. Attrition rates from clinical trials are often 30 percent or higher for older medications such as the TCAs and around 15 percent for newer options such as the SSRIs.

The newer psychotherapies appear to do as well as medication during the acute depression phase, although the number of studies is fewer and the findings are not always consistent. One typical study found that IPT alone was about as effective as medication alone (with each better than a control condition) and that the combination was better still. In general, medication relieved symptoms more quickly, but IPT produced more improvement in social functioning and quality of relationships. The combined treatment retained the independent benefits of each.

IPT also fared well in the 1989 National Institute of Mental Health Treatment of Depression Collaborative Research Program. The TDCRP, as it is known, is perhaps the most influential study to date that compared medication and psychotherapy. In that trial, patients with major depression were randomly assigned to 16 weeks of IPT, CBT or the TCA imipramine, combined with meetings with a psychiatrist or a placebo plus meetings. Patients with less severe depression improved equally across conditions. Among more severely depressed patients, imipramine worked faster than IPT, but both were comparable by the end of treatment and both were superior to a placebo.

As for CBT, most of the published trials have found it to be as effective as medication in the acute phase. The most notable exception--the TDCRP--did find that cognitive therapy was less efficacious than either medication or IPT (and no better than a placebo) in the treatment of more severely depressed patients. Because the study was large and was the first major comparison to include a pill-placebo control, its results considerably dampened enthusiasm for cognitive therapy, even though no other study had produced such a negative finding.

Today this conclusion appears to have been premature. More recent studies have found that CBT is superior to pill-placebos and is as good as an SSRI for more severely depressed outpatients. These studies suggest that cognitive therapy's success depends greatly on the level of a therapist's training and experience with it, especially for patients with more serious or complicated symptoms.

Continuing the Fight
The best treatments for reducing acute distress also seem to work as well for reducing relapse when they are carried into the continuation phase. Antidepressants appear to reduce the risk for relapse by at least half. It is unclear exactly how long patients must keep taking medication to pass from remission into full recovery, but current convention is to go for at least six to nine months.

IPT during the continuation phase appears to prevent relapse nearly as well as medication, although studies in this regard are few. Recent investigations also suggest that if cognitive therapy is continued past the point of remission, it can reduce the risk for relapse. To date, no studies have compared continuation CBT to continuation IPT or medication.

During the maintenance phase, medication is usually recommended for high-risk patients, especially those with multiple prior episodes. Therapy can go on for years. It does protect against recurrence. Even among recovered patients, though, the risk of recurrence off medication is at least two to three times greater. Given that there is no evidence that prior medication use does anything to reduce subsequent risk for recurrence, most physicians will encourage their high-risk patients to stay on medication indefinitely.

Studies of maintenance IPT are few, but they generally support the notion that it, too, reduces risk of recurrence. It has not been as efficacious as keeping people on medication, but the handful of studies have typically cut back the frequency of IPT to monthly sessions while maintaining medication at full, acute-treatment dosages. It would be interesting to see how maintenance IPT compares when the psychotherapy sessions are also kept at "full strength."

Several studies have shown that CBT has an enduring protective benefit that extends beyond the end of treatment. Patients treated to remission with CBT were only about half as likely to relapse after treatment termination as patients treated to remission with medication, and the CBT patients were no more likely to relapse than patients who continued on the prescriptions. CBT appears to produce this enduring effect regardless of whether it is provided alone or in combination with medication during acute treatment and even if it is added only after medication has reduced acute symptoms. Further, indications are that this enduring effect may even prevent wholly new episodes (recurrence), although findings are still far from conclusive.

Given these trends, CBT may ultimately prove more cost-effective than medication. Psychotherapy usually costs at least twice as much as medication over the first several months, but if the enduring effect of CBT truly extends over time, it may prove less costly for patients to learn the skills involved and discontinue treatment than to stay on medication indefinitely. It remains unclear whether other interventions such as IPT have an enduring effect, but this possibility should certainly be explored.

Our review of the treatment literature indicates that some forms of psychotherapy can work as well as medication in alleviating acute distress. IPT may enhance the breadth of response, and CBT may enhance its stability. Combined treatment, though more costly, appears to retain the advantages of each approach. Good medical care can be hard to find, and the psychotherapies that have garnered the most empirical support are still not widely practiced. Nevertheless, some kind of treatment is almost always better than none for a person facing depression. The real tragedy is that even as alternatives expand, too few people seek help.