30 Years of Listening to Prozac: A Discussion with Peter Kramer

Peter D. Kramer, MD, is a psychiatrist and the author of eight books, including the influential 1993 book Listening to Prozac (now available in a new 30th anniversary edition) and, most recently, the novel Death of the Great Man. His essays, op-eds, and book reviews have appeared in the New York Times, Wall Street Journal, Washington Post, and elsewhere. For almost forty years, he practiced psychiatry in Providence, Rhode Island, where he is an emeritus professor of psychiatry and human behavior at Brown University. He now writes full-time.

Awais Aftab, MD, is a clinical assistant professor of psychiatry at Case Western Reserve University. He is interested in conceptual and philosophical issues in psychiatry and is the author of this Substack newsletter.

---

Aftab: I’d like to congratulate you on the 30th anniversary of Listening to Prozac. It’s a remarkable book with an enduring cultural and professional impact, and one that still holds up quite well despite the passage of time. I really enjoyed reading your new introduction to the 2023 edition and found your reflections quite illuminating.

You remark on how psychiatric practice is very different today. The way we assess patients, the time we spend with them, the questions we ask, the diagnostic labels we use, and perhaps most importantly, the sorts of observations we make. I remember that when I first read Listening to Prozac, what struck me the most was the way you practiced. My immediate thoughts were, “How can I learn to practice like Peter Kramer! Why didn’t they teach this in psychiatry residency!” It seems to me that psychiatrists are no longer trained in a manner that allows them to recognize and articulate the changes you observed in your patients. If SSRIs were only just now entering the market, would psychiatrists of my generation even notice the changes in temperament? And if someone did notice, would those observations remain a clinical curiosity, at the level of anecdote?

Kramer: Thank you, Awais, and thank you especially for the comment about my approach to patient care and also about the evergreen aspects of the book.

When Penguin invited me to revisit Listening to Prozac for a thirtieth anniversary edition, I returned to the text with apprehension. It was a surprise and a pleasure to discover that, with few exceptions, I could stand by what I’d written in the early 1990s. My impression is that the book’s premise has held up well: that antidepressants affect temperament and that the brain’s handling of serotonin is somehow involved. Of course, the continued relevance of observations from three decades back does not necessarily speak well for the field’s recent progress in research, particularly on the translational side.

As to how I practiced: I came to medicine via literature and philosophy, and, so, psychoanalysis. In medical school, at Harvard, two mentors pulled me away from strict Freudianism, Milton Mazer and Leston Havens. Milton drew me into community psychiatry and a pragmatic approach to patients grounded in social psychology. Les introduced me to existential psychotherapy and, more, encouraged me to take an interest in a range of theories.

In the early 1970s, when I trained, Harvard was psychotherapy-centered in the extreme. We were expected to treat psychotic patients using variants of psychoanalysis. If medication was introduced, it was handled by psychopharmacologists, who were subspecialists. I stumbled on their work only late in med school, not via the core curriculum. This sidelining of medication may have prolonged patients’ suffering, but as a trainee, you learned to listen attentively and to monitor your own feelings in session.

By the time I’d entered private practice, in the 1980s, I’d been exposed to neurobiology and medical model psychiatry. In that new introduction to Listening to Prozac, I explain how, in the Carter Administration, through happenstance I came to work under Gerry Klerman in a Health and Human Services position that familiarized me with the entire federal research portfolio in mental health and substance abuse. But my work with patients was still centered on psychotherapy.

Later, when I first prescribed SSRIs, it was attention to something like countertransference that caused me to realize that they were acting like co-therapists. The relief that the medications afforded created a framework that allowed the therapy to proceed. And patients told me what the drugs were doing.

I can’t say that I know just what clinical psychiatry looks like now, but for many years, at Brown, I taught a segment of the basic psychotherapy course for psychiatry residents. As the medical model and more routinized approaches to psychotherapy took hold in training, the residents often appeared resentful. I was teaching them a painstaking method, or many methods—we covered five or six therapeutic schools—, and they did not have time to apply them. I was asking too much.

But, yes, I think that there’s a danger that if you make diagnosis central to treatment, and you accept that sign and symptom checklist entries determine diagnosis, then when you prescribe medication, your focus will be constrained. It will be hard to notice change that occurs outside the framework. I erred in the other direction. I was aware of diagnosis, but I did not rely on it heavily. I wanted, as Leston Havens taught, metaphorically to sit beside patients, view the world as they did, and experience some of what they were experiencing. When patients told me that, on medication, they were more assertive and socially adept, I knew that they were right, knew it experientially.

Kramer: There’s a danger that if you make diagnosis central to treatment, and you accept that sign and symptom checklist entries determine diagnosis, then when you prescribe medication, your focus will be constrained. It will be hard to notice change that occurs outside the framework.

Aftab: You talk about how “neurosis” was a catchall diagnosis for psychiatrists of your generation, covering all but the most severe disorders. You write at one point in the new introduction:

“Hearing a recitation of a scroll-length list of ailments—depression, bipolar II, post-traumatic stress disorder, attention deficits, and insomnia—I would listen and think: tough childhood, demanding job, bad marriage. With some patients, I would feel nostalgia for the neurosis umbrella. The core problem, in my dated view, was a tendency to undermine the self... It's good that doctors are hearing patients' complaints, but sessions are brief and the encouragement to subdivide seems to obviate the need to integrate—to identify an overarching problem and try addressing it first.”

There is a move towards, or I suppose a return to, umbrella diagnoses in psychopathology. I am thinking of dimensional models such as HiTOP, in which multiple comorbid DSM diagnoses can be viewed through the unifying lens of higher-order spectra such as the “internalizing spectrum.” Internalizing spectrum may very well be the descriptive analogue of neurosis, and antidepressants do seem to show efficacy across the internalizing spectrum. What if antidepressants, when they were marketed, had been marketed as “anti-neurotics”? Would our popular understanding of these agents be any different today?

Kramer: A thought-provoking question. Different labels lead to different narratives.

Prozac arrived just as depression was becoming discussable. William Styron’s memoir Darkness Visible was published shortly after. It made depression out to be a devastating illness that can strike anyone and for which transformative treatments are few. In that environment, being an antidepressant legitimated Prozac as a medicine for a serious ailment. My account, that SSRIs did some of the work of psychotherapy, took place within that framework. So did patient care. People with depression were treated under the aegis of medicine, and, if they were good responders, they just happened to do better overall, socially and psychologically, in ways that went beyond the indication for prescribing.

Thirty-five years ago, would a pill for neurosis have sounded legitimate? Imagine marketing (when his persona had a clearer valence) an anti-Woody-Allen drug. Or an alpha-status enhancer. Or psychotherapy in a caplet. If pharma got the medications past the FDA (doubtful), it’s not clear whether they’d have seen wild sales figures or a big bust. The premise might have been out ahead of what the culture was prepared to accept.

But let’s imagine, as I think you’re proposing, that antineurotic drugs found a role and were later incidentally discovered to be solidly effective for depression. The sequence might have shaped the connotations of depression, with a risk that it would be understood as a mere consequence of oversensitivity and emotional fragility—weakness in temperament. Removing the value judgment, some depression may be that way, grounded in neuroticism and adversity, but much is not.

As for the eventual acceptance of the drugs, I find it possible to imagine that from any starting point, the SSRIs would have made their way. Recent alarmism notwithstanding, they’re plenty likeable. People are on them for a reason.

Regarding the identity confusion that leads to your question—at base, what are these meds?—it may be that we are inching toward an understanding or resolution. In the new introduction and afterword, I make note of bodies of research with diverse starting points, clinical and neurochemical, that see SSRIs, and perhaps all of the older antidepressants, as subserving two separable functions, restoring resilience and altering emotional bias. By this account, the action on emotional bias combats neuroticism, while the combination allows for recovery from depression. We can imagine future drugs that would make a pharmacological dissection, by affecting only one or the other element. But then, these sorts of matters are rarely resolved cleanly. Nor would it surprise me if down the road, researchers discovered that these drugs have a direct action on depression, not mediated by resilience—that they are antidepressant and antineurotic both.

Kramer: In the new introduction and afterword, I make note of bodies of research with diverse starting points, clinical and neurochemical, that see SSRIs, and perhaps all of the older antidepressants, as subserving two separable functions, restoring resilience and altering emotional bias. By this account, the action on emotional bias combats neuroticism, while the combination allows for recovery from depression.

Aftab: When you wrote Listening to Prozac, it was not common for patients to be maintained on antidepressants. Most patients stayed on antidepressants for about 9 – 12 months. Do you think psychiatry’s embrace of maintenance treatment as the default option for most patients was justified in retrospect? Would we benefit from a return to the older model of time limited treatment of acute episodes?

Kramer: Regarding relapse, I’m slightly suspicious of the research or, more, the use of imperfect research results to encourage an insistence on knocking down all symptoms and continuing medication, or increasing doses or adding drugs, until every marker of depression has disappeared. Drug companies benefit, but do patients? I like to keep medication regimens simple and leave some of the mop-up work to psychotherapy.

In Ordinarily Well, I describe the approach to duration that I used until 2018, when I closed my practice. Even treating chronic or recurrent depression, I tried regularly to cut back on medication doses and, if symptoms did not worsen, finally to discontinue. I’m thinking of reliable patients in weekly psychotherapy—the norm, for me—or patients seen less often who knew, or whose family members knew, to contact me in case of trouble. I would see emergent episodes, but I caught them early, and I’d say that I had good results overall, with few bad recurrences.

A number of patients did seem to need full maintenance doses. Even in that group, I rarely saw bad side effects. Perhaps I was treating a favorable population, but the way that antidepressants are portrayed online bears no relationship to what I encountered in the office.

I tapered meds very slowly, but I also had patients who, when they felt free of depression, simply came off medication. They forgot to refill a prescription and forgot to tell me that they had forgotten—every doctor will be familiar with the pattern. Those patients mostly continued to do well.

This question would benefit from a comparison trial—maintenance versus taper-and-rescue. It would need to be done over the course of years, to allow for slow discontinuation, and of course you would have the problem of patients signing on and later changing their minds about whether they were comfortable with the regimen they were assigned to. Complex to evaluate statistically, I fear, as there would be so many patterns. So: we may never have an objective guide to treatment.

Meanwhile, the short answer is: Yes, my impression is that the field has gone too far in the direction of maintenance as the default approach. I’m not confident here. Often enough, in practice, I was drawn to relying on long-term maintenance treatment.

An afterthought—about my lack of confidence in the other direction. Researchers’ implicit faith in a narrow version of evidence-based medicine has set the academy against practitioners and raised suspicions that what doctors arrive at through daily practice is often misguided. I mostly begin with the opposite assumption, that despite all the research on sources of bias, clinicians are good observers, and that we do well to listen to their extrapolations about cause and effect. But on this question, maintenance, I sometimes lean the other way. Much drug tapering is too rapid, abrupt discontinuation can lead to intolerable effects for patients, and practitioners may confuse withdrawal with recurrent illness—and, so, favor maintenance on faulty grounds. And that’s without considering the more extreme hypotheses that have medications worsening outcomes in the long term. So out in the field, it may be hard to get a clear view of which treatment approach is best.

Kramer: Researchers’ implicit faith in a narrow version of evidence-based medicine has set the academy against practitioners and raised suspicions that what doctors arrive at through daily practice is often misguided. I mostly begin with the opposite assumption, that despite all the research on sources of bias, clinicians are good observers, and that we do well to listen to their extrapolations about cause and effect.

Aftab: I often wonder why so many people stay on antidepressants. I am not satisfied with the standard logic that people stay on antidepressants to reduce the risk of future depressive episodes. It does apply to some people, but I think there’s more going on. There’s a group of people who seem to derive active symptom benefit from antidepressants. Unlike episodic depression, these patients have dysthymia, or some other chronic mix of anxiety and depressive symptoms (like chronic pain) and taking the antidepressant isn’t really about preventing some future episode but rather about alleviating symptoms that are always there. Then there’s another group of people who experience the sort of temperamental changes you describe, and they stay on the meds because they find those personality changes to be beneficial. (And perhaps there are also some patients who stay on antidepressants because they try to taper too quickly and experience withdrawal.) Clinical trials have focused mostly on relapse prevention, ignoring these other reasons why patients stay on medications. At the same time, we seem to have lost the language to express this. Depression has become this large, bloated category. Even patients who experience changes in temperament may interpret and describe it as “the medication is working for my depression.” What are your thoughts on why people stay on antidepressants?

Kramer: As you know, early in my experience with SSRIs, I found that some patients reported feeling and functioning better globally on medication than off, and in ways that often translated into a better quality of life, better success with career or family. I’d guess that this dimension does explain some patients’ preference for staying on meds. I’m not sure that diagnosis needs to be invoked. People find that they’re better off, and they’re reluctant to risk returning to times when they did less well.

Aftab: In many ways, the issues that you identified in Listening to Prozac remain underexplored by the scientific community. Mainstream psychiatry has largely ignored the question of what it might mean if we had the power to tweak personality in ways that patients find desirable or that the culture rewards. Commentators who were interested in this question in the 1990s appear to have lost interest in the ethics of cosmetic psychopharmacology once skepticism about antidepressant efficacy became mainstream. The idea that these medications that barely beat placebo in clinical trials could produce such pronounced effects on personality became too implausible for many to entertain seriously. You speculated in Listening to Prozac that we were entering an era in which awareness of medication effects would shape how we understood ourselves. We certainly entered an era of mass antidepressant use, but I’m not sure that this shaped how we understand ourselves collectively in any meaningful way. From my perspective, our common and routine use of antidepressants is mostly devoid of the curiosity needed to ask the questions that you were asking.

Kramer: When I coined the term “cosmetic psychopharmacology,” I was imagining a future in which psychopharmacologists became progressively better at tweaking personality in narrow and predictable ways, and that future never came. In effect, as research made it ever more plausible that SSRIs were reshaping temperament, they became the best example of the phenomenon—the closest we’ve gotten. It’s not surprising that interest in the topic waned. Philosophers had said what they had to say, and no dramatic new examples came down the pike.

As a side note, I wish that researchers had remained curious about antidepressants’ effects on temperament. Perhaps the field will regain interest once we have access to a range of new antidepressant treatments. Do they all color personality in the same way, or are there results that are particular to SSRIs?

Kramer: I wish that researchers had remained curious about antidepressants’ effects on temperament.

On the broader issue, I think that the widespread use of antidepressants has contributed to the medicalization of depression, and perhaps of temperament as well. Today, it’s commonplace for parents to say jokingly that they were neutral on the nature and nurture question until they had children—because they’re all born different. In the Freudian era, it was the impact of subtle differences in environment that was obvious. The “refrigerator mother” theory of autism was plausible because, via the importance of the Oedipus complex, personality and mental illness were assumed and, consequently, observed to be shaped by subtle emotional injuries in early development.

I agree that much of the American public now doubts the efficacy of antidepressants—I argue in detail against this conclusion in my last nonfiction book, Ordinarily Well: The Case for Antidepressants—or has become blasé about these drugs. They bore us. At the same time, the public may also believe that antidepressants have personality effects. (“If you take my husband off that pill, you’re going to have to put me on it!”) More generally, views of the mind as basically “biological” have become implicit in our thought and discourse. The integration of the various assumptions—the adoption of a medical model, but in the context of mistrust of science and of psychiatry in particular—has led to a lack of enthusiasm for any standard treatment, whether medication or psychotherapy, and sometimes to a desperate turning toward dramatic and unproven remedies.

Aftab: You write: “A culture that lacked the concept mental illness would still recognize depression as a multisystem disease, like diabetes.” The heterogeneity of our notion of depression gives me pause. It is the case that there are subsets of people with depression who show more pronounced endocrine, inflammatory, and metabolic changes, but there are many depressed individuals who don’t show these changes, or show these changes to a degree that would otherwise be considered normal. And it’s also the case that these changes are not entirely specific to depression either. Some people with other psychiatric disorders also show these changes to varying degrees. So while I accept that depression seems to be linked to adverse changes in multiple systems, I can’t quite bring myself to call it “a multisystem disease, like diabetes” because that seems to reify depression in a manner not supported by validators.

Kramer: Your question takes us beyond Listening to Prozac and into the territory of Against Depression. I’m not sure that, in 2025, I’d be up for designing a twentieth-anniversary edition. Too much work! And there’s been too little improvement in our understanding of what depression is, biologically, at its core or, as some would have it, cores.

Your objections are good ones. I don’t want to dismiss them. But in hopes of holding on to the “multisystem” concept a bit longer, allow me to think aloud—to play with the concept—and address a portion of what fails to persuade you.

Let’s take Type 1 diabetes as the exemplar, multisystem disease in purest form. Type 1 involves problems with cells in the pancreas. The result is a failure of insulin production and, so, hyperglycemia and subsequent problems with vision, hearing, sexual function, blood circulation, heart and kidney health, and so on. What would it take to disqualify type 1 diabetes—to pull it out of the multisystem-disease category?

Latency is not the problem. Years ago, with fewer tools to manage glycemic control, patients would tend to suffer from complications about fourteen years after the onset.

Nor is uncertainty about the mechanism of damage disqualifying. For most of my career, there was debate over the role of glycemic control. Other factors seemed to be involved. For all this time, type 1 diabetes was a multisystem disease. The attribution is independent of any understanding of how the pancreas links to, say, the kidney.

Direct damage to multiple organs does not need to be central to the ailment. With type 1 diabetes, the pathology is in the pancreas, and complications are due to indirect action.

Is it a convincing objection to this formulation—multisystem—to say that other disorders can have the same effect? Type 2 diabetes can, and gestational diabetes, and diabetes arising from rare genetic variants, some affecting insulin production and, some, cell receptors. We call these disorders diabetes, but in another culture, each might have a distinct name. Even pre-diabetes—what might be called non-illness—puts the kidneys, heart and nerves at increased risk. You can get the multisystem pattern of diabetes in the absence of the full disease—but no matter.

Nor does it disrupt our understanding of type 1 diabetes if unrelated autoimmune disorders, ones that act primarily on cartilage or blood vessels, can give rise to a generally comparable collection of downstream ailments.

Type 1 diabetes is itself a collection of diseases variously arising from infection or autoimmune problems, so there’s not even a single entity “type 1 diabetes.” Heterogeneity of forms and causes is not disqualifying. Nor is disparity in outcomes. Some patients in bad control fare well for years, and some in good control do poorly. Variability is the norm in multisystem diseases.

Lifestyle choices—behaviors—can alter the course of downstream outcomes in diabetes. That fact, too, does nothing to shake our confidence in our categorization.

You can see where this argument is heading. Let’s choose as our core form of depression the type that arises from genetic liability, early adversity, and later exposure to stressors—depression where the episodes in time become longer, more complex, and more closely spaced, with relapses eventually recurring almost independent of circumstance. Let’s stipulate that this core depression leads variably to a host of complications in diverse organs, blood and bone and heart and endocrine glands and brain, likely via stress hormones but perhaps from other hormonal or immune dysregulation or for reasons not yet under consideration. As with diabetes years ago, we have a key villain in mind, but we’re far from sure.

If we accept tentatively that our core depression is a multisystem disease, we will not be dissuaded by the fact that the downstream organ damage is variable, indirect, and delayed in time, nor by the fact that other disorders can damage the same organs. Yes, we see similar harm in depression arising from endocrine or circulatory disorders or traumatic brain injury (and a host of other posited causes like inflammation or problems with gut flora). Subclinical variants may be damaging, too. With schizophrenia or PTSD, the correspondences are less precise. In no case will those observations be disqualifying.

If we resist this formulation—depression as multisystem disease—will it be because what’s at issue is a mental illness, defined, at its core, by problems with the mind—hopelessness and low or flat mood? But whenever we ask whether a mental illness is a multisystem disease, the starting point will be along those lines.

Kramer: If we resist this formulation—depression as multisystem disease—will it be because what’s at issue is a mental illness, defined, at its core, by problems with the mind?

To pull the comparison together, let’s undertake a thought experiment. Imagine a civilization on Mars and a type of Martian in whom diabetes, instead of presenting as dehydration, lassitude, blurred vision, and sweet urine, presents instead as a mental disorder—that’s how doctors there know that diabetes is present—and say also that Martian scientists have not discovered the Islets of Langerhans or elucidated the role of glycemic control. Although Martians with the disorder may suffer from the mental illness, this Martian diabetes mostly does not kill them immediately. Instead, within fifteen or twenty years, the afflicted will tend variably to develop organ damage throughout the body. Is Martian diabetes a multisystem disease?

I do not present this argument seriously. Obvious objections come to mind. Often, the core function of analogy is to shine light on differences. But the simple concept, depression as multisystem disease can do some good work, and although it has shortcomings, I don’t find it easy to dismiss.

Let me now make a point on the other side—also not to be taken too seriously. I think that most people, even most doctors, ignore the frequency with which everything correlates with everything. If you have one or two disorders, you’re likelier to have any other disorder. It’s as if there were such a thing as “bad protoplasm” or widespread effects of exposure to stress, adversity, environmental toxins, malnutrition, and other broad-spectrum insults to mind and body. Put differently, the healthier and luckier you are, the likelier you are to stay healthy and lucky. We should not be overly impressed with correlations. They may not be telling us much that is specific to the disorders under study.

Aftab: In the 2023 edition, you consider the possibility that SSRIs have two distinct and independent effects: they diminish neuroticism (via serotonergic pathways) and increase neuroplasticity and counter depression (via effects on BDNF and TrkB, etc.), and some patients experience both these effects simultaneously in a synergistic manner. How confident should we be about this possibility?

Kramer: Not confident. As I say in the current afterword to Listening to Prozac, the TrkB findings make no sense. Why should widely diverse compounds developed to affect the brain’s use of serotonin (or norepinephrine and so on) all happen to work instead via their binding to TrkB? I find the theory attractive because it links so well to research on emotional bias. I’m thinking of the work of Philip Cowen, Catherine Harmer, and their colleagues at Oxford. Well, and also, the TrkB part of the story is buttressed by impressive research. But I do not find the new theory especially plausible.

Aftab: You briefly discuss the antipsychiatry critiques of antidepressants in the new edition. What stands out to me the most about some critics of antidepressants is that they proudly come from the tradition of “evidence-based medicine.” Some skeptics in the EBM movement have a distinct disregard for clinical experience and lived experience, and only seem to care about what randomized, double-blind clinical trials have to say about medication efficacy. Evidence-based medicine and antipsychiatry are strange bedfellows at face value. How do you make sense of this intersection?

Kramer: The devil can quote scripture.

The intent is to hoist the profession on its own petard. These guys and gals are having fun.

Some of the undermining begins with researchers’ creating experiments or designing meta-analyses meant to show treatments in their worst light. Some is simpler cherry—picking of unfavorable data. A few arguments draw on zombie data, that is, results that have already been disproved. (In Ordinarily Well, I examine an egregious case involving antidepressants and apparent placebo effects.) Some is misdirection, like the recent attempt to discredit antidepressants through casting doubt on the links between brain serotonin status and depression.

To be fair, there are good reasons for skepticism about the research in a certain era that was sponsored by drug companies. And publication bias is a serious problem. There’s plenty to work with. But many of the articles grounded in that work are simply tendentious.

Here we’re in the territory of my most recent nonfiction book, Ordinarily Well. It’s true that there, I write that I find it gratifying when the results of horserace research support observations from clinical experience. But my more substantive complaint is that the drug-skeptics focus too narrowly on commercial drug trials. In the process, they ignore other systematic outcome research and, more importantly, miss any sense of how the body of knowledge adds up. In my view, the notion that antidepressants have no direct main effect on depressive episodes is untenable. I rely on research findings to justify that conclusion. But it’s true that in the background, my experience with patients influences my version of the synthesis.

Kramer: My more substantive complaint is that the drug-skeptics focus too narrowly on commercial drug trials. In the process, they ignore other systematic outcome research and, more importantly, miss any sense of how the body of knowledge adds up. In my view, the notion that antidepressants have no direct main effect on depressive episodes is untenable.

Ideological opponents of psychiatry don the mantle of EBM because research data is the coin of the realm. Often, their arguments are what, in Moments of Engagement, I called scientomorphic—the term may have been coined by the poet Vicki Hearne—that is, they need to have the form, if not the content of science. But at base, I think that antipsychiatry arises from a world view or set of preferences that is at odds with contemporary science, a nostalgic, rear-guard position in which emotional difficulties arise almost exclusively from experience and should be addressed almost exclusively with remedies that are experiential.

As to what the experiential treatment is, the objectors to medication are divided. Some favor exercise, meditation, or yoga—not treatments to which EBM has been especially well applied. Some favor highly systematized psychotherapy although historically, the antipsychiatry movement mistrusted systems.

The reliance on a narrow version of EBM is often selective or asymmetrical. When discussing harm from antidepressants, opponents of medication tend to accept reports based on clinical and lived experience.

Aftab: You write: “The prospect of competition is good news for SSRIs. For decades, they have carried a heavy load as the treatment of choice for almost all depression. Imagine internists trying to treat all high blood pressure with only one sort of pill. It’s preposterous. If a range of effective treatments emerge, SSRIs will find their place.” What do you suspect their place will be if ketamine, psychedelics, and neuromodulation such as accelerated transcranial magnetic stimulation also establish their value as first-line agents of depression? What role do you think psychotherapy should have in the treatment of depression?

Kramer: SSRIs will look more consistently effective if researchers identify subgroups of patients likely to be especially responsive. New, competing treatments should help stimulate that research. It would be gratifying if clinical observations guided the studies—if practitioners happened to notice that a given medication or electromagnetic intervention works best for a certain type of patient and researchers tested the observations.

Clinician observation can be a powerful tool. I recall a highly favorable study of nefazodone, a drug that then fell by the wayside in part because practicing doctors did not find it to be especially effective. My guess is that the crowdsourced result was right.

Since I am not a research scientist, I prefer for my comments to be grounded in clinical experience, and I have none with psychedelics. I don’t see that they have been studied especially well. They have not been tested much in the absence of psychotherapy. In some trials, there’s idiosyncrasy in the patient recruitment—you get patients who already like psychedelics. Often, psychedelics are tested on patients still taking approved antidepressants. And there’s that amusing study of ketamine in patients undergoing general anesthesia—I write about it in the afterword. But the word of mouth is favorable. I’m all but certain that ketamine and psychedelics will find a role in depression treatment.

I’m curious about the allopregnanolone drugs. Supposedly a single long infusion or a two-week course of capsules puts patients in good shape for months, so they’re not chronically on medication.

But for all the complaints, SSRIs and SNRIs may prove hard to beat. I wouldn’t be surprised to find that they still have a role for a long while—say, for the remainder of my lifetime.

As for psychotherapy, yes, yes, yes. I spent almost all my clinical hours, forty years’ worth, practicing psychotherapy. In patients who were not gravely ill, my ideal was to use medication adjunctively, if at all.

I am skeptical of research that shows progress in psychotherapy, except in the sense that psychotherapy must speak the language of its time. A treatment built around awareness of penis envy would not do well today. We demand attention to the trauma narrative. But my bet is always on what used to be called nonspecific effects, common to most treatments—caring, empathy, wisdom, insight, common sense, the doctor-patient match, and so on. Practitioners whose strengths are on those lines may have trouble adhering to protocol.

But I have always liked psychotherapy. My third book, Should You Leave?, is built around a collection of quirky midcentury psychotherapies that I found intriguing.

Kramer: As for psychotherapy, yes, yes, yes. I spent almost all my clinical hours, forty years’ worth, practicing psychotherapy. In patients who were not gravely ill, my ideal was to use medication adjunctively, if at all.

As a practitioner, I always relied on psychotherapy—psychotherapy first, when possible, and psychotherapy aided by medication when called for. I don’t know how clinicians do without it. I love the title of Leston Haven’s book A Safe Place. In early meetings especially, you need for patients to feel secure and to shed some of the symptom overlay arising from the catastrophic thoughts that led them to seek help. Sometimes, it’s only after the early drama has passed that you get a sense of what’s chronic—the symptoms that should lead you to formulate the DSM diagnosis. Of course, the reverse can be true. It can take a few sessions before the patient trusts (or resents) you enough to reveal the more extreme symptoms that should inform the diagnosis.

Aftab: Dr Kramer, thank you!

---

This post is part of a series featuring in-depth interviews and discussions intended to foster a re-examination of philosophical and scientific debates in the psy-sciences. See prior discussions with Diane O’Leary, Richard Gipps, David Mordecai, Emily Deans, Nicole Rust, Rob Wipond, and Martin Plöderl.

---