Ordinarily Well: The Case for Antidepressants


Better than Well

WRITING ABOUT PROZAC and personality, I coined two phrases, better than well and cosmetic psychopharmacology.

The first theme came from a patient who had prided himself on his cynicism and irascibility. He had said that on Prozac he was “better than baseline,” more comfortable and less prickly than he had been even before his episode of depression had begun. He was ambivalent about his newfound equanimity, but in conventional terms (and his wife’s) it represented change to the good. When other patients offered similar reports, their focus was assertiveness. Relief from depression aside, on medication they were more confident and outgoing.

The second idea came from an ethical and practical challenge I had faced. A patient would respond to Prozac in this fashion—better than well—and then come off medication. Months later, still mostly free of depression, she might return to my office. She had functioned more effectively while on Prozac. Would I offer it again?

This sequence made me wonder: Might a doctor prescribe for a person who had never been ill, to provide a boost in social ease? I called this potential use of medication cosmetic psychopharmacology, on the model of cosmetic surgery. Both efforts employed medical technology to move a person from a normal state to another normal state that is better rewarded in the culture. The discussion in Listening to Prozac energized the study of what ethicists call enhancement, medical interventions used for purposes other than the cure or prevention of disease.

Better than well referred to enhancement in psychiatric patients. Cosmetic psychopharmacology referred to enhancement in people with no mental disorder.

I first outlined these concepts in essays (“Metamorphosis” and “The New You”) addressed to colleagues, saying that they must have seen what I had.

Later, in an interview, Jonathan Cole recalled that, treating patients with Prozac before the drug came to market, he had noticed some who “were clearly better than they had ever been before in their lives,” adding that “there were just enough of them to make a difference.” Cole called me “the Listening to Prozac man.” Later, he retrieved my name: “Peter Kramer can be somewhat foggy but he makes valid points.” Oh, dear. All the same, Cole confirmed the better-than-well phenomenon.

Coles’s conversation was with David Healy, a noted critic of drug companies and arguably our most important historian of psychopharmacology. Healy asked the Swedish neuroscientist Arvid Carlsson—he would later win the Nobel Prize in Physiology or Medicine—about cosmetic psychopharmacology. Carlsson ventured, “I think that is true now with [Prozac] and all these drugs. There are people who feel so much better, who didn’t have any diagnosis really.”

This expert testimony spoke to the concepts’ plausibility.

So did prior research. One motivation for developing serotonin-specific antidepressants had been the hope that they would influence phenomena beyond mainstream depression. These included obsessive-compulsive disorder, impulsive aggression, and atypical depression, where patients eat and sleep too much—and often suffer from social anxiety and low self-worth. In different ways, serotonin regulation had been implicated in all these problems. When I noticed personality effects in response to Prozac, I had reason to suspect that I might be onto something.

Was I? After Listening to Prozac appeared, researchers began to address cosmetic psychopharmacology and the better-than-well phenomenon.

The earliest studies asked whether healthy people experience personality change on medication. To mention one trial from a handful: David Healy looked at twenty normal subjects, including members of his research team at the University of Wales. They took either Zoloft or reboxetine, a medicine (not marketed in the States) that alters the way the brain handles norepinephrine. Later, participants who had been on one medicine were given the other.

Some volunteers did well on Zoloft, others on reboxetine. Unfavorable reactions to each drug also occurred. People who were suited to Zoloft reported feeling less prone to emotional upset when on it. In contrast, those who responded to reboxetine might experience unpleasant agitation on Zoloft.

When participants responded favorably to Zoloft, they became more sociable. Overall, the authors reported, “On their preferred drug, subjects scored lower on aggression, agitation, hostility, psychasthenia [low-level obsessions] and somatic anxiety, while scoring higher on social desirability.”

Healy stressed a point that we considered when discussing how attention to average outcomes understates antidepressants’ worth: on a given drug, some patients fare well and some especially poorly.

More recently, neuroscientists interested in ethics tested whether the SSRI Celexa could make thirty normal volunteers “prosocial.” The subjects were paired off and put in a game where, with money at stake, they could make choices that would spare others harm, but at a cost of unfairness to themselves. On the SSRI, more than when on placebo or an antidepressant that affects norepinephrine, the participants behaved collaboratively, avoiding choices that would injure other people.

Because the issue is not clinical—few if any doctors are medicating “squeaky-clean” people, with no hint of depression—cosmetic psychopharmacology has been of interest mainly to philosophers. No one can say with certainty whether SSRIs induce self-assurance in healthy people. But what research there is suggests that personality effects are common and that on SSRIs people gain social comfort.

Funding is more readily available for studies of illness. Some trials of fair size have looked at the better-than-well phenomenon, personality change in depressed patients.

In 1999, Lisa Ekselius and other Swedish researchers examined more than three hundred moderately depressed patients drawn from primary-care clinics. In the trial, participants were given an SSRI (Celexa or Zoloft—with no risk of getting placebo), in treatment overseen by their general practitioners. This design has advantages. Depressed people will volunteer since all participants receive effective treatment. The doctors know the patients well. We are in Roland Kuhn territory, avoiding distortions that arise in trials that use placebos.

In Ekselius’s hands, Celexa and Zoloft were outstanding treatments. In six months, 75 to 80 percent of patients who entered the trial and over 90 percent of those who completed it improved substantially on measures of depression. These figures resemble the ones Kuhn had reported four decades before.

(We should bookmark this result so that we can return to it when we discuss placebo effects: Do we believe that virtually all depression seen in general medical practice would respond to dummy pills?)

The better-than-well question is, When depressed people improve on an SSRI, do they experience personality change as well? On medication, the Swedish patients became less anxious and less aggressive. They scored higher on measures of socialization and social desirability. The study tracked many traits—detachment, suspicion, obsessiveness, and others. The authors observed, “After treatment, significant changes in the direction of normalisation were seen in all scales.” Personality shifts occurred even in the few patients who did not get less depressed. Statistically the gains in sociability had almost no correlation with change in mood.

These results—reliable relief from depression and, independently, normalization of maladaptive personality traits—go a long way toward explaining the enthusiasm for these medications.

Others replicated the Ekselius experiment. UCLA researchers found decreased “harm avoidance” (that is, increased boldness) in patients with obsessive-compulsive disorder who improved on Paxil. Evidently, you did not need to start out depressed to experience personality effects from medication. Controlled trials told a similar story. In 2000, a team at Beth Israel Hospital in New York studied more than four hundred patients with dysthymia—chronic low-level depression. Those who recovered on Zoloft showed the greatest lessening of harm avoidance, with patients on imipramine and placebo trailing.

What causes what? Do people naturally become less timid and more sociable as they emerge from depression? A Finnish study suggested as much, but it was not a controlled trial, and its design might have led it to miss medication effects. Conversely, do antidepressants bring about personality change that facilitates recovery from mood disorder? Considering this question, researchers have focused on a trait called neuroticism. The coinage, from Hans Eysenck, encapsulates negative thinking, uncomfortable self-consciousness, and emotional vulnerability and instability. A 2008 Canadian study found that SSRIs work via their ability to mute neuroticism. When it drops, then depressive symptoms do, too.

Fortunately, the field has a painstaking randomized trial on this topic. It comes from a prolific group at the University of Pennsylvania and Vanderbilt University, largely psychologists interested in psychotherapy. In 2009, the Penn-Vanderbilt team reported on 240 depressed patients treated with Paxil, cognitive therapy, or a placebo pill.

For depression, both psychotherapy and Paxil outperformed placebo. But with medication especially, the greater change was in muting neuroticism. On Paxil, “patients reported changes in neuroticism and extraversion that were 4 to 8 times as large as the changes reported by placebo patients.” On medication, patients gained confidence, social ease, and emotional stability.

Placebo did not diminish neuroticism. Paxil did, and the effect was consequential. After an initial four-month trial, the team followed the patients for another year. Patients whose neuroticism scores had dropped suffered fewer depressive episodes. Even patients who stopped treatment retained protection—so long as they had first improved on Paxil. Patients whose neuroticism fell during cognitive therapy remained liable to recurrent depression. In the researchers’ words, “Perhaps the most surprising pattern we observed was the relation between neuroticism reduction during acute SSRI treatment and subsequent resistance to relapse.”

What Paxil did best was to make people less fragile, the effect that my patients had discussed with me. This change proved more stabilizing than anything that psychotherapy or placebo use did.

Paxil, the Penn-Vanderbilt team wrote, “appears to have a specific pharmacological effect on personality that is distinct from its effect on depression.” The results, they said, “support the notion that SSRIs’ effects on personality go beyond and perhaps contribute to their antidepressant effects.”

The studies on SSRIs suggest that my main error was in caution—in my having written that “on occasion” medication caused personality change. On antidepressants, patients feel and become more emotionally secure much of the time. In the Penn-Vanderbilt study, when Paxil (contrasted to placebo) was considered a treatment for neuroticism, the effect size was 0.6, in the high-medium range. Today, SSRIs’ better-than-well effect is less controversial—less contradicted in the literature—than the notion that SSRIs constitute a reliable treatment for depression. That is also to say that the antidepressant debate is possible only because conventional rating scales miss the drugs’ influence on quality of life and traits such as neuroticism.

Because of these findings and, more, because of doctors’ own observations in the course of patient care, medical practice has changed. Twenty years ago, if an episode of depression ended, everyone was satisfied. But having seen the shift in temperament that can accompany antidepressant use, doctors began to worry over patients who, although technically free of depression, remained diffident, socially withdrawn, emotionally brittle, and insecure. Whether to prescribe in these cases is a difficult question.

In Listening to Prozac, I warned about “diagnostic bracket creep,” the tendency of disease categories to expand to embrace what medication does. The Penn-Vanderbilt findings blur that issue. Perhaps attention to personality traits constitutes basic care—prevention of recurrence—and without any change in our definition of illness. For depressive patients, neurotic traits no longer seem benign. In the minds of some doctors, the state I called better than well is a facet of ordinary health.

Often, in discussions of the antidepressant controversy, I am challenged about the better than well coinage: How’s that working for ya? The idea is that when there is doubt about medications’ ability to perform their main job, treating depression, it is ludicrous to think that they might do more. But the issues, interrupting depressive episodes and reshaping personality, are distinct. While the clamor against antidepressants has been in crescendo, the better-than-well hypothesis has received increasing confirmation and proved increasingly relevant to clinical care.

If this attention to my thought has been gratifying (it has), it also represents a common sequence in medical practice. Work with patients inspires certain ideas, a doctor floats them, colleagues confirm the observations, patient care shifts, and along the way research offers further support. Today, we are so focused on error that we may imagine that folly is the rule in medical practice. My impression is different. Clinical wisdom tends to pan out, if in general terms. Often, systematic trials suggest refinements, experience adds further pointers, and so on. The back-and-forth takes place in the context of current scientific knowledge—the context of biological plausibility. This process, progress through a virtuous dialectic between less and more formal observation, is the norm. Grand revelations of error, although important when they occur, are the exception.

If you find an error or have any questions, please email us at admin@doctorlib.info. Thank you!