DESPITE THE NEW research tools, antidepressants gained acceptance slowly. To give a sense of the pace of change, I want to share a personal history that a colleague, Robert Liberman, has generously made public.
I have mentioned that by 1965 imipramine had been tested on a thousand patients and shown substantial efficacy. That analysis appeared in a review compiled by two men I came to know later in their careers. Jonathan Cole ran the first major trials of psychotherapeutic drugs in the United States. He was an affable raconteur, at once a top-notch scientist and a low-key ambassador for pharmacology. His younger colleague Gerald Klerman, although by no means humorless, had a sterner brilliance. He would become something like an enforcer for evidence-based psychiatry.
Considering everything from chemical structure to side effects, in their review Cole and Klerman showcased imipramine as a medication that science could characterize in detail, a treatment that had come of age. But they were respectful of objections, among them the argument that “what clinical efficacy these drugs have is mediated by socio-psychological mechanisms, particularly the suggestibility and faith of the patient, the enthusiasm amid zeal of the physician, or both.” Cole and Klerman traced the claim for classic placebo effects to a 1961 paper, “A Criticism of Drug Therapy in Psychiatry,” which appeared in a publication of the American Medical Association. Liberman, then a second-year medical student at Dartmouth, was the author.
Regarding imipramine, Liberman wrote that it had been “greeted with much undeserved praise as an antidepressive agent.” His approach to the evidence was hardly evenhanded. He highlighted unsuccessful treatments reported by a private practitioner while underplaying a systematic study, summarized in the same journal, that found favorable outcomes. In support of the idea that the drug works via placebo effects, Liberman cited an article that did not discuss depression.
Liberman’s paper reflects the scholarship of its time, but it serves now mainly to show how far back the antidepressant-as-placebo argument extends and how little was required to sustain it. In concluding that prescribing amounted to “poor medical practice,” Liberman championed mainstream views, widely held by psychotherapists and welcome in leading publications.
When I located the essay, it occurred to me that I knew a Robert Liberman, an advocate for medication use in depression. I telephoned, and he acknowledged the critique as his own, adding that it had a bittersweet follow-on, one that he had discussed in published interviews.
Leaving Dartmouth, Liberman had transferred to Johns Hopkins, where he became depressed. In retrospect, he had suffered episodes since age sixteen. At Hopkins—just after his “criticism” appeared—Liberman was offered an antidepressant. He recovered quickly. He next became depressed while working at the National Institute of Mental Health. When psychotherapy did little for him, he took an antidepressant, to good effect. Still, he was unconvinced.
For another episode, in California in the 1970s, Liberman again chose psychotherapy. It brought no relief. Years later, discussing that course of treatment, he said, “Despite the fact that I did not benefit from their ministrations and even got worse, these psychiatrists never suggested to me that I might improve with antidepressant medication.”
Liberman’s own doubts persisted. Only after another twenty years of recurrent illness, requiring hospitalizations and finally electroconvulsive therapy, did he embrace antidepressant use. By then, in the mid-1990s, psychiatry had adopted the idea of a midway state between recurrent major depression and bipolar disorder, a variant in which patients never quite become manic but suffer worsening depression over time. Looking at the alternation between Liberman’s periods of depression and intervals of productivity, a psychopharmacologist made that diagnosis, although the possibility of straightforward recurrent depression remained in play.
The midway condition can respond poorly to antidepressants—but the consultant had Liberman’s favorable past experience to rely on. Ever since, Liberman has been on an antidepressant. Despite his disorder’s progressive nature, on medication Liberman has gone fifteen years without a substantial depressive episode. He has been healthier—less burdened by depression—than in his earlier years.
Liberman has integrated his experience into his clinical practice. When he considers openness appropriate, he tells patients that he takes medication daily, a self-revelation that encourages others to adhere to their own drug regimens.
Do we censure those California psychiatrists, the ones who did not prescribe and left their patient on a downhill course? We must be careful not to judge them according to our current awareness of progressive risk to brain and mind. Likely, Liberman’s caregivers believed—this position still has its advocates—that without gains in self-understanding, cure does more harm than good.
Still, there is something undoctorly about withholding medication from an ailing patient who has responded to it repeatedly. Not anecdote but ideology is the antithesis of science. Anecdotally, antidepressants worked for Liberman.
Even a skeptic focused on placebo effects might have prescribed for a patient in Liberman’s circumstances, but nowhere in medicine do placebos act in the way that antidepressants did for Liberman, consistently interrupting complex manifestations of illness and conferring solid stabilization. Psychiatrists in my generation saw this sequence regularly: psychotherapy failed for patients who believed in it stubbornly, and then antidepressants succeeded. Those cases were convincing.
To return to Cole and Klerman’s early review of research on imipramine: The paper was evenhanded in its presentation of the placebo issue that Liberman had raised in 1961. Imipramine has evident side effects. It can make patients so dry-mouthed that they or an interviewer—hearing the tongue cluck as it releases from the roof of the mouth—might guess that they were on medication. The trial would no longer be blind.
Critics argued that antidepressants might appear less effective if tested against “active” placebos—otherwise inert drugs that cause side effects such as dry mouth and constipation and, so, inspire higher expectancy. On this topic, Cole and Klerman found few relevant studies. In one, participants on a symptom-causing placebo fared especially poorly. Had the constipating pill done harm? The jury was out on whether active placebos had a role to play in testing. (To this day, it remains out. There have been few new studies. What you believe depends on how you handle the outlying data—and what you make of the powers of placebo pills in general, a topic we will confront in time.)
Cole and Klerman were likewise frank about the limits to scientific knowledge. In developing antidepressants, drug companies screened compounds that alter the way that nerve cells handle certain chemical messengers in the brain, substances (such as norepinephrine and serotonin) called monoamines. The idea that they matter in depression was the monoamine hypothesis. Cole and Klerman cautioned, “The evidence for this view is almost entirely indirect.” Even the best-supported parts were “admittedly speculative.” Drugs that influenced norepinephrine and the rest often did offer relief, but it was never clear that monoamine abnormalities caused or constituted depression.
I mention Cole and Klerman’s circumspection because of the role that the monoamine hypothesis plays in the current antidepressant controversy. Critics complain that faith in medication is sustained by an unsubstantiated belief that drugs reverse depression by boosting transmitters whose deficiency constitutes the disorder. (Marcia Angell writes that the theory has been “broadly accepted … by the medical profession.”) Particular aspects of the monoamine hypothesis have gained plausibility in recent years—serotonin regulation does have something to do with depression—but for the whole of the time that antidepressants have been available, leading figures in psychiatry, including the pioneering psychopharmacologists, have viewed the theory skeptically. In my training, it was presented as, to use the language I employ in Listening to Prozac, “perhaps false and at least incomplete.”
The prevailing understanding is not that antidepressants reverse a fundamental deficiency (of serotonin or a related chemical), but rather that the medications restore resilience in the mind and brain, allowing the growth of new nerve cells and the elaboration of new connections between cells. The drugs “permit” depression to diminish by allowing repair and new learning to proceed in brains and persons previously left “stuck” in depression. This theory of recovery has driven research since before Prozac’s time.
In the antidepressant controversy, the monoamine hypothesis has become a red herring. In each generation, acceptance of antidepressants has been sustained by complex accounts of the science like Cole and Klerman’s.
Another constant, from the early going, has been empirical support from randomized controlled trials. Cole and Klerman entertained a host of caveats, but finally their review attested to imipramine’s efficacy: a two-to-one advantage, drug over dummy pills, in the treatment of depression—and the summary numbers were conservative. Some trials had used the drug at low doses. When patients received at least 200 milligrams daily, the results might be stronger.
Discussing randomization, I referred to Archie Cochrane, the guiding spirit for today’s curators of evidence-based medicine, the Cochrane Collaboration. Cochrane was a Scottish doctor who practiced in prisoner-of-war camps in World War II, went on to study lung disease, and made his name as an early proponent of randomized trials.
Cochrane, who expressed grave doubts about psychotherapy, was critical of psychiatric research in general. But in 1972, in his influential book on effective health care, when he listed treatments that might improve outcomes for the British National Health Service, Cochrane put antidepressants on a short list with blood-pressure pills, the polio vaccine, cortisone, and antibiotics for tuberculosis.
Cochrane’s endorsement and Liberman’s troubling history epitomize conflicting attitudes toward depression treatment in the 1970s. Expert researchers considered antidepressants highly effective, but many clinicians refused even to test out medication and remained doubters.