Robert Spitzer on DSM-III: A Recently Recovered Interview


In this hour-long interview, recorded by invitation at his home outside New York on February 22, 2006, Robert Spitzer, Task Force Chair to two major updates to the Diagnostic and Statistical Manual of Mental Disorders, discusses at length his rationale for adding more than 100 new disorders, including 7 new anxiety disorders, to DSM-III in 1980.

Robert Spitzer
Robert Spitzer

The Columbia University psychiatrist volunteers how critics might have derailed these and other changes; his decisions over inclusion, exclusion, expansion, and renaming; the controversies that flared over psychoanalysis and psychodynamic approaches; the relation of homosexuality to normal development; the risk of false positives and of pathologizing normal reactions to difficult situations; and the role that drug companies played in sponsoring key research and conferences used to justify some of the new disorders.

Long believed to have been lost, the interview is published in full for the first time after it was recovered in a recent move.

The interview below has been edited for length and clarity, to avoid repetition.
Christopher Lane: Several articles on you mention that before you developed an interest in biometrics, your career actually began with the psychoanalytic theories of Wilhelm Reich. I’m curious, how did that come about?

Robert Spitzer: I think when I was nine or ten my mother sent me to a psychoanalyst because I had slapped her. But I didn’t really have therapy. When I was a 15 year old kid, I had an English teacher who had had a training with Reich. He was a student of Reich and did Reichian therapy. I was intrigued by his approach. I was a radical of sorts so it appealed to me.

Lane: With the diagnostic work you helped to develop after the Washington University group in St. Louis and its focus on “consistent and reliable” rules for diagnosis, what happened to your clinical interest?

Spitzer: I had psychoanalytic training and graduated from Columbia—graduated but barely. I guess I was a little bit of a troublemaker. When I was a research fellow I did some psychoanalysis. But I never really felt very comfortable doing it. I was unsure if I was helping anyone and often didn’t know what I was doing.

Lane: I’d like to focus this afternoon on a couple of related issues: the diagnostic issue of validity and its central importance to DSM-III. And the classification of new anxiety disorders in 1980 such as generalized anxiety disorder, social phobia (later renamed social anxiety disorder), and panic disorder. In an article co-authored with Ronald Bayer on the immediate history of DSM-III, you wrote that the Task Force you chaired for several years “had its intellectual roots in St. Louis instead of Vienna. And derived its intellectual integration from Kraepelin, not Freud.” Given your training at Columbia, were you concerned about discrepancies between the two approaches—that in pursuing Kraepelin, say, the unconscious dimensions of symptom-formation would fall out of the picture?

Spitzer: I think the way we saw it … The Kraepelin approach just meant we were interested in description. We were interested in the course of the illness and we avoided theories about ideology.

Lane: We can come back to the question of ideology, but from a treatment perspective does that not ignore or downplay another set of concerns about the meaning of symptoms and the underlying issues that may be said to produce them?

Spitzer: Well, sure it did. It left them all out! From a psychoanalytic perspective, if you look at something like panic disorder, there’s a drive and there’s something that’s unconsciously warded off. We felt, ‘That was an interesting theory, but we wouldn’t know what to do with that.’ So …

The whole approach we took with DSM-III was to develop a diagnostic system that people who had totally different orientations could still use. So if a psychoanalyst thought that panic disorder was based on impulses that were warded off, he or she could still use this system. Behavior therapists too … If you’re asking if it’s a different approach: Sure, it’s a different approach.

Lane: How did you personally negotiate the tensions between these approaches?

Spitzer: I negotiated the discrepancy by withdrawing my cathexis from psychoanalysis. I mean, I suppose it could have been that by day I was a DSM [sic] and by night I was a psychoanalyst, but it didn’t work out that way, partly for intellectual reasons and partly for personal ones. My experience with psychoanalysis—and I’ve had many different attempts at it—have largely been unsuccessful. So …

For the first few years while I began working on DSM in 1974, I was still seeing patients until that stopped. I used to make a joke that my private practice became increasingly private until there was no one there [Laughs]. I think the maximum I ever did was 15 hours per week.

Lane: And moving between these different worlds?

Spitzer: Yes, they’re very different worlds. I think a lot of my success in DSM was being able to negotiate with different groups. You know, having an analytic training certainly enabled me to deal with people better, although … it became a real conflict at the time.

Lane: You’ve drawn mock-serious comparisons between the peace treaty negotiated between Egypt and Israel, which would put the stakes high indeed. What was driving analysts’ frustration with the DSM process?

Spitzer: I think they may have felt betrayed. Well, the first betrayal was with the homosexuality thing [its de-listing as a mental disorder and removal from the DSM in 1973, a development Spitzer helped to negotiate]. Particularly the analytic group at Columbia felt very betrayed and angry with me for that.

Lane: Their position at the time was so retrograde and non-psychoanalytic—even quite anti-Freudian …

Spitzer: You mean that Freud didn’t regard homosexuality as an illness? True. But he also didn’t regard it as part of normal development. Well I don’t regard it as normal development.

Lane: By the end of his career, surely Freud had thoroughly upended the idea of normal development …

Spitzer: [Not apparently hearing.] You know, I actually wrote to Reich once and he wrote me back. When I did my experiments and they all came out negative, I wrote to ask why I wasn’t getting good results. This must have been in 1952, I’d guess. And I got back a reply from him saying the reason was because of the atomic bomb testing. [Laughs.] Unfortunately I didn’t keep the letter—it would’ve looked good in a frame.

Lane: Let’s go back to the key issue of pathologizing and de-pathologizing. Taking you further back to 1968, if I may, with the publication of DSM-II, you co-published with Paul Wilson “A Guide to the APA’s New Diagnostic Nomenclature.” The article is of great interest to me because in it you discuss “the elimination of the word ‘reaction’ from labels such as ‘schizophrenic reaction,’ ‘paranoid reaction,’” and so on—a major development, surely, in how we conceptualize and describe psychiatric diagnosis. Was there lengthy discussion at the time about making that change?

Spitzer: About doing that? No, there was no discussion at all. No, no. You have to understand: the APA had decided with DSM-II to use the ICD-8. The ICD-8 was written by one person, [Sir] Aubrey Lewis at the Maudsley [Institute of Psychiatry, London], and he didn’t have the word reaction so, for us, there was never any discussion.

Lane: In hindsight, as the DSM-III task force ended up arguing over commas, parentheses, and quite a bit more, surely that was a major redefinition …

Spitzer: Well, the DSM-II adopted ICD-8 with just some very minor, little changes. You could make a comparison. There were some subtypes and conversion disorders—I forget what they were … very trivial changes.

Lane: But deleting the word reaction from quite a few designated kinds of mental illness—in a diagnostic manual that’s also meant to define them and for clinicians to recognize them—is still a major shift because it’s altering the ontological status of the condition …

Spitzer: Yes. Yes, it is a major shift. I think if there had been any discussion, it would on the order of ‘We don’t add anything by just putting the word reaction to everything.’ You can still believe in psycho-biology without having the word [reaction] there. That would have been the argument. But I doubt there was any argument, because by that time, just having the word reaction didn’t mean very much.

Lane: Except that removing it meant you were in effect turning a reaction to something into more like a lasting, possibly lifelong state. One without an obvious cause, in that you also removed stressors that might be tied to environment, economic status, family dynamics, and so on …

Spitzer: Well, what we were saying is ‘Dropping the word reaction doesn’t really mean anything.’ I think that’s probably true—I don’t think it did mean very much. With DSM-III there were huge controversies over this and other developments when it came out. But with DSM-II, I guess there was one article, possibly in a newspaper, where William Menninger suggested that by adopting [European-based] ICD-8 we’re losing the contribution of American psychiatry. Now whether he was responding to the reaction thing I don’t recall. I know there was that one complaint.

Lane: I’ve read different accounts of how the DSM-III Task Force started and how you were appointed, given your background and work on the previous edition. If you look back now on months like April 1979 when the issue about anxiety neurosis, already around for a while, clearly …

Spitzer: … yes, it came to a head then …

Lane: How would you characterize those months now?

Spitzer: I don’t remember them well. I remember the final assembly meeting where they gave me a standing ovation. Because at that point the analytic group was really trying to defeat the whole thing. And there was a possibility that they might be successful. It was all over dysthymia versus neurotic depression. Also, my marriage was falling apart at the same time. So there was a lot going on.

Lane: If we can turn to a November 1975 article you co-wrote with Endicott and Robins, “Clinical Criteria for Psychiatric Diagnosis in DSM-III,” your clearly stated rationale was to reduce variability while also increasing reliability in psychiatric diagnosis. The article nonetheless discusses variables that complicate those ends, such as ‘subject variants’—presumably, how an illness can take different courses and flare up at different times—and an ongoing interpretive issue about whether two or more diagnosers really see things the same way. Both go to the heart of interrater reliability.

Spitzer: A patient says something: how do you know that it meets …? Two people may say, ‘Well that fulfills that criteria’ and somebody says, ‘No it doesn’t.’ That’s a big problem. Still is a big problem.

Lane: As far as I can tell from the correspondence over DSM-III, your perspective at the time was …

Spitzer: … that the variance that goes into diagnostic unreliability would be minimized.

Lane: Or possibly even solved for good …

Spitzer: Well, it’s still there. We’ve been accused of exaggerating how much reliability improved. I don’t know if you’re familiar with the work of Herb Kutchins and Stuart Kirk. They’ve written several books critical of DSM-III and -IV. And their argument was that we overstated the improvement in reliability. As part of that they quote Gerald Klerman [Spitzer’s colleague], who in a debate on DSM-III said ‘The problem of [diagnostic] reliability has been solved.’ Which is regretful, because it hasn’t been solved at all. I think there’s some truth …. If you ask clinicians now how reliable is the DSM or how much did it improve it, I don’t know what they would say but it’s a modest improvement.

It also depends on the settings. In specialized settings, like if you go to an eating disorder clinic or an anxiety disorder clinic, you can—particularly if you use structure carefully—get very good reliability, 0.8-type. But if you just do patients in an outpatient setting, you’re talking 0.6, which is … very poor—very modest.

Lane: Does that lead you to rethink any of the focus on Kraepelin?

Spitzer: I don’t know what you are going to rethink. Where are you going to go? [Laughs.]

Lane: Maybe rethink some of his emphasis on categories and observable behavior? After all, the whole impetus was to minimize their unreliability.

Spitzer: What are you going to do if you find you can’t minimize that to the extent that you’d like? I mean, what do you do at that point? Is there another approach that you’re going to take? There was an article in the New York Times about a month ago about this new psychoanalytic diagnostic system. There’s an analytic group that have produced a psychodynamic diagnostic system that’s supposed to be a complement to the DSM, not in competition with it … It’s an interesting approach, to get away from the observable behavior. How successful they’ll be I don’t know.

Lane: In a way this links back to the St. Louis group, where we started. In his 1972 essay “Diagnostic Criteria for Use in Psychiatric Research,” Feighner acknowledges: “Consistent and reliable laboratory findings have not yet been demonstrated in the more common psychiatric disorders.”

Spitzer: That’s for sure.

Lane: Is that still the case, after quite a few DSM updates?

Spitzer: Oh I think it’s absolutely the case. It’s interesting, in talking about DSM-5, over which I have no role at all, some people in the APA have said, “What we’ll try to do in DSM-5 is bring biological etiology into it,” and I think that’s just silly and premature. There’s no way to do that at the best of times.

Lane: Only biological etiology, not that of psychology or social determinants …?

Spitzer: No, that’s not what they were talking about.

Lane: With Feighner, I went to his criteria for anxiety neurosis [included in DSM-II]. Of course he doesn’t have subtypes for the diagnosis, but he’s very precise about its thresholds. He says “there must have been at least 6 anxiety attacks, each separated by at least 1 week from the others,” and so on. As the various subtypes and forms of anxiety disorder were made the focus of your task force, did you believe you could keep to the same degree of precision and consistency or come to think Feighner was setting too high a bar for diagnosis to occur?

Spitzer: Simply I’ve forgotten the Feighner criteria for anxiety. I think it was really just panic, given his mention of “attacks.” So he didn’t recognize Generalized Anxiety Disorder [first listed in DSM-III 8 years later, in 1980]. Well, we came up with that name after we had anxiety neurosis in DSM-II and if you had panic there had to be something that was left over, so that became generalized anxiety disorder. And then there was discussion of social phobias. Well, there was always the notion that there were phobias. So then, the common notion of phobias was specific phobias. And then people said, ‘Well but some phobias are more generalized,’ and that became generalized phobia. And later that became social phobia. And now with DSM-IV it’s social anxiety disorder, which is a better term than social phobia.

Lane: Can we return to the moment when generalized anxiety disorder and panic disorder were starting to be differentiated, because that’s surely quite a complicated step.

Spitzer: Well phobias don’t generally lead to panic attacks. So I don’t think there was much of a discussion of separating panic disorder from the phobias.

Lane: I was referring to your saying that GAD had to be distinguished from panic disorder because “there had to be something that was left over …”

Spitzer: Well, you see, we started with anxiety neurosis, DSM-II. Along comes Don Klein [a colleague at Columbia] and says there’s this thing called panic disorder, which Feighner had called anxiety neurosis. If we’re going to recognize panic disorder, our thinking was, clearly the DSM-II anxiety neurosis was a broader category, so we need a name for that other part, and that was generalized anxiety disorder.

Lane: Right. But the thing that’s so complicated about social phobia—a phobia of people, a diverse, immensely varied object—is that it doesn’t seem to work like a conventional phobia, which I guess is why it was renamed …

Spitzer: No, it’s not a conventional phobia.

Lane: So how did it get distinguished from specific phobia, which has a narrower target?

Spitzer: Well, I think, you know, with specific phobia there are things that scare people and they avoid them: heights, tunnels, snakes, dogs, things like that. And then people said, ‘Well, some people avoid people, so let’s call that social phobia. I mean, let’s call that generalized phobia.

Lane: OK, but do you see how separating them out gets so tricky? I mean, there was also avoidant personality disorder [similarly introduced by DSM-III in 1980] …

Spitzer: Well, avoidant personality disorder [laughing] is basically identical to social phobia. That developed from the personality people, who wanted to have avoidant personality disorder, and really it was embarrassing that it was clear that it was the same thing as social phobia.

Lane: I was discussing this recently with Michael Liebowitz and Richard Heimberg, who I think were on the DSM-IV Anxiety Disorders group. Their preference was to think about social phobia as a continuum with chronic and milder forms, with subtypes that they came up with, based on situational anxieties such as public-speaking anxiety and awkwardness over eating alone in a restaurant. Clearly I’m an outsider to this process, but when I see these and other disorders so clearly proliferating, with their subtypes so obviously overlapping, yet dual diagnosis is not just permitted but often encouraged, there are clear signs of redundancy and a major risk of overdiagnosis. One really needs to go back to an earlier moment to see how these things got set up and why they seemed so crucial at the time—when people like Feighner seemed fine about representing anxiety neurosis on its own, as a generic category for all kinds.

Spitzer: Well, concerning my memory, I remember very clearly how generalized anxiety disorder came about. My memory of specific phobias shifting to social phobias is not quite as clear. Essentially, it always came down to people on the Working Group, and did they recognize clinically a subgroup that they thought ought to have a different name.

Lane: So they were coming up with criteria that they believed they could operationalize with you—that was the way to initiate discussion around inclusion?

Spitzer: Right.

Lane: Let me quote this interesting account of the Task Force, which goes back to the start of the interview: “The Task Force believed that the large body of etiological evidence put forth by those committed to a psychodynamic perspective could not serve as the basis for defining the diagnostic categories.” We’ve already covered some of this, and I know you wanted consistency, a multiaxial approach, and an a-theoretical one. It’s also clear from reading alternative perspectives on this process, including by the psychodynamic community, that they really felt that they did have a set of diagnostic categories to offer. It’s also the case there is always some etiology in DSM-III—it’s not completely missing and thus a-theoretical—even in how things are named, such as phobias and disorders.

Spitzer: Well, the only etiology is one by definition. You know, panic disorders or distress disorders. Post-traumatic stress disorder: you could say the etiology is there. Adjustment disorder. But aside from those disorders, where by definition there’s a stressor, there’s no etiology.

Lane: Presumably also, social phobia …

Spitzer: Right. You know, the analytic critique of DSM-III as it was developing was not, ‘We have another way of classifying.’ It’s not like they said, ‘Here’s our system for categories.’ Their main complaint was that in the description of the disorders we left out psychodynamic factors.

Now there’s an interesting story about that. The American Psychoanalytic Association had a liaison committee that met with me. And the chair of the committee said, ‘You know, the problem is, we know so much more about these disorders than DSM-III puts in there. Well, what he meant was the psychodynamic [side]. So I said, ‘Why don’t you take one disorder and write it up the way you think, with the stuff you think we don’t have in there.’ So they gave it to Larry Rockland, who happened to be a friend of mine. He took OCD and he wrote up the psychodynamics. And it was embarrassing. I mean he had included ‘anal conflicts’ and stuff. And so I said, ‘This just won’t fly. This is not something we can use.’

[CL seeks to clarify, 2022: Rockland’s proposal made no such characterization. He instead wrote Spitzer: ‘It seems to be part of the general argument throughout DSM III that lists of symptoms are somehow more scientific and ‘harder’ facts than muddle-headed psychodynamic theorizing and fantasizing. I think that this is a very unfair characterization of psychological thinking.’ Rockland added that he wanted more attention paid to ‘the particular shades and varieties of intrapsychic conflict, which exists in all psychopathology and in all people.’ Hence his ‘attempt at a positive diagnosis of anxiety disorder based on a psychodynamic understanding of the patient, the patient’s conflicts, and the psychosocial stresses operative which add up to a coherent picture of why the patient developed an anxiety neurosis at this time.’ Rockland to Spitzer, “Some Thoughts on the Subject: Should Psychodynamics Be Included in the DSM III,” unpublished memo c. January 1978, quoted with APA permission in Lane, Shyness: How Normal Behavior Became a Sickness, p. 54.]

Lane: There’s an interesting paradox there in that when psychoanalysis tries to be precise diagnostically or to establish strict inclusion and exclusion criteria, people immediately rebel against it. It may be something about the approach itself that needs to be less schematic or formulaic.

Spitzer: Have you had an analytic training yourself?

Lane: No, though I’ve contemplated it many times. I have good ties to the Chicago Psychoanalytic Institute and worked with a psychoanalytic theorist all through graduate school [Jacqueline Rose], focusing mostly on the history of psychoanalysis and psychiatry. From that perspective, this would seem to be a recurring issue for the field about definition and reductionism, though it’s also a dynamic in the culture at large, with managed care and the insurance companies wanting limited criteria to reimburse for any kind of psychotherapy.

Spitzer: Well, the whole … There’s a very interesting book called The Fall of an Icon: Psychoanalysis and Academic Psychiatry by Joel Paris [2005]. It’s a history of the rise and fall of psychoanalysis within American academic psychiatry. It’s really very good. And it’s not anti-analytic. It’s critical, but says the reason psychoanalysis declined was because it didn’t produce efficacy data. It’s not a critique of analysis itself.

With the liaison committee, they weren’t making criteria. They were making etiological psychodynamic formulations.

Lane: It’s also my understanding from the DSM correspondence that psychoanalysts John and William Frosch were brought in following the liaison committee to help represent the psychodynamic side, though that was surely quite a few years after things had already gotten well-underway …

Spitzer: Well it was first John, and then John quit …

Lane: Yes. There are some details about this in the Bayer article, with John Frosch requesting that anxiety neurosis stay in DSM-III and that ‘an effort be made to define descriptive frames of references for neurosis before dangling the term …’ That’s from the article. ‘Nothing came of this proposal,’ Bayer continues, ‘although one-and-a-half years later the argument was more successful.’ I think at that point there was the ‘Neurotic Peace Treaty,’ you had called it, with some kind of compromise.

Spitzer: Gee, I totally forgot that. So Frosch had suggested ‘Don’t give up neurosis, let’s try to come up with a set of frames of references’? I don’t know what that would have meant. It’s interesting that European psychiatry also used the term ‘neurosis.’ It’s in ICD. So when I said, ‘In DSM-III we’ve got to get rid of the word ‘neurosis’ because it has psychoanalytic meaning,’ the answer to that would have been, ‘Not necessarily,’ given the international classification. So we could have developed criteria for neurosis that would not be psychodynamic. But it would have been very difficult.

Don Klein made the point that the problem with neurosis was not the exclusion criteria, but the inclusion ones. In other words, you didn’t know whether to include, say, eating disorders. Is that neurosis or not? Should you include somatoform disorders?

Lane: The term doesn’t perhaps rule out enough …

Spitzer: Right.

Lane: But you just said—I’m struck by it—that the task force needed ‘to get rid of the word because it has psychoanalytic meaning.’ Rather than, say, because it has unclear etiology or poor rule-out criteria.

Spitzer: I meant that we could have made an attempt to keep the term ‘neurosis,’ but it would have been …. I mean I don’t know what the boundary would have been. We could have kept the traditional DSM and had, you know, Dissociative Mood Disorder. But it would have made … We wanted to keep all the mood disorders together, whether they were psychotic or not. So what would you do? You’d have to have Neurotic Depression, Psychotic Depression, and that distinction had kind of fallen by the wayside.

Lane: So nobody said, ‘Well look if you turn to ICD-8 you’ll see in fact that the Europeans kept the term …’

Spitzer: No, nobody did that.

Lane: It would have been interesting if they had. What would you have done?

Spitzer: Well, it would have been a tough argument. I would have said, ‘It would be hard to define what the boundaries of that category would be.’ Now it’s true that the Europeans use it and they don’t mean unconscious conflict. I suppose the reason they use it is from the history of Freud and the categories he recognized and developed. Freud didn’t recognize eating disorders. If he had, he might have considered it a neurosis.

Lane: Possibly. There are various references to not eating in his case studies, and so on, but yes, not as full-blown disorders.

Spitzer: But seriously, Bill Frosch was the only Task Force member who raised the issue of ‘Should we have diagnostic criteria in DSM-III?’ He said we should have two manuals: one for researchers, which would have diagnostic criteria, and the other not, for clinicians. And we rejected that. We wanted to have a similar system. But seriously, the one psychoanalytic member of the Task Force …!

Lane: So he stayed on to work with Bill …?

Spitzer: No, Bill replaced John. I don’t know why John left. I don’t know if he left because he felt we weren’t utilizing his talents or maybe he had other things to do. But he left.

[CL in 2022: His cryptic letter of resignation is reproduced in Shyness, p. 60: “I respect the efforts and the energy of the participants but I cannot help wondering at times ————.”]

Lane: But then, again, I guess he too didn’t quite manage to get …

Spitzer: …. the psychoanalytic community onboard?

Lane: No, the psychoanalytic and psychodynamic perspectives represented in the manual …

Spitzer: No, no. First of all, he was only one person. I don’t think they ever expected him to do very much. He was pretty much a token figure.

Lane: Hmm. A few more questions, if I may. I wanted to ask about a couple of diagnoses you described as promising but that weren’t ultimately included in DSM-III because the criteria weren’t “developed-enough.” One of them was “Late Luteal Phase Dysphoric Disorder,” ultimately renamed PMDD — Premenstrual Dysphoric Disorder.

Spitzer: Well, by the time DSM-IIIR came along [7 years later, in 1987], we did propose it and it ended up in the Appendix.

Lane: I’d like to ask about this wider pattern—the process whereby some disorders were ruled out, such as Introverted Personality Disorder, but others seemed sufficiently encouraging to you to include or put in the Appendix for possible future inclusion. Were these decisions tied to concern about false positives?

Spitzer: [Laughing] Well, I don’t know if Introverted got any further than a letter from me. I don’t recall any discussion where … I would write all these letters back to people, and if there was some point that I thought I was sympathetic or I thought a lot of other people were, I would bring it to them, if it was just an idea. You know some Jungians said, ‘You ought to have introverted personality disorder,’ and I thought it was not a promising idea. So the Personality Committee would never hear of it. Or I might send them a letter, I don’t know.

We didn’t really think very much about false positives. Are you familiar with Jerry Wakefield and his critique? He’s written “Disorder as Harmful Dysfunction: A Conceptual Critique of DSM,” “Defining Mental Disorder,” and others. He’s very concerned with false positives and thinks the DSM’s definition of mental disorder is too broad and true of a lot of situational disorders that are essentially a normal reaction to that context. So he’s very concerned with false positives but it’s not something we were at all.

Lane: There was some discussion, too, about the role of pharmaceutical companies in helping to push and to popularize the newer disorders included in DSM-III—for instance Upjohn, maker of Xanax, over the inclusion of panic disorder. Isaac Marks told me recently about a key conference in Boston on panic that Upjohn had sponsored that was on trying to “operationalize” its criteria.

Spitzer: What did he say?

Lane: He said the CEO had gotten up to speak and stated there were three reasons why Upjohn was there, taking an interest in the issue and its potential DSM inclusion: the first was money. The second was money. And the third was money. Marks was marveling that they were so blatant about it—that they didn’t even try to mask it.

Spitzer: What happened: during the development of DSM, Upjohn was marketing and doing studies on Xanax. There was one committee meeting where they paid for—I think they donated—a place, a site where we met as a convention, and they paid for it, which was a mistake on our parts. We should never have done that.

But they had no influence on any criteria or the name. So this thing that we were influenced by pharmaceuticals is something that I just, I say, is just absurd. There was never any discussion. They never made any attempt, you know, ‘We would like you to define it more broadly’ or anything. It never happened.

They were delighted that we had the category panic disorder, because they felt they had a drug for it. But they never tried to influence how the criteria were developed or anything.

Lane: I would note several articles in the press detailing very high percentages of DSM committee members also receiving pharma dollars for their research. When I also see, say, GlaxoSmithKline quoting various Key Opinion Leaders in psychiatry on social anxiety disorder—for instance Murray Stein, also a pharma recipient—saying there’s a continuum in severity from anywhere from 2 percent to around 18 percent of a population, almost one-in-five, depending on the settings, GSK will always advertise: ‘Experts such as Dr. Stein say as many as 18 percent may be affected.’ In that way, a disorder that was to a large degree clinically invisible a few years earlier, because not in the DSM, all of the sudden becomes the third-most diagnosed disorder in the U.S., after depression and alcoholism. That raises many flags about DSM inclusion, as well as about your Task Force’s decisions to include, and how, and what conflicts of interest were operating.

Spitzer: Are you saying pharmaceutical companies have a tendency to magnify the prevalence of disorders for which they have a treatment? Well, sure. There’s another interesting phenomenon: researchers also give maximal prevalence to the disorders they have a particular interest in. In other words, if you’re really interested in panic disorder, you’re going to tend to say it’s very common. You never hear an expert say, ‘My disorder is very rare.’ Never. They always tend to see it as more common.

Lane: On the other hand, it’s tough when the cultural ramifications of such maximalism are extensive …

Spitzer: But I don’t see how that’s a DSM problem.

Lane: Well, to the extent that the inclusion of, say, Late Luteal Dysphoric Disorder in the DSM Appendix presents opportunities for greater research …

Spitzer: Actually, that’s the only category in the Appendix for which the FDA has approved a treatment [Sarafem/Prozac], which is kind of interesting. Before, the FDA would only approve treatments for recognized disorders.

Lane: So, how did that happen?

Spitzer: Well, there was enough research evidence suggesting that it was a valid category, which I think it is. So I guess Eli Lilly [maker of Prozac and Sarafem] convinced them that even though it’s only in the Appendix, uh, we’ve got a drug for it. I mean the only reason it’s only in the Appendix is feminist opposition to it. Otherwise … And it’ll be in DSM-5 or something like that.

Lane: Could we close with DSM-5 and the future trajectory of the field? What other categories and conditions do you think are likely to be included, or should be?

Spitzer: Well, I think PMDD will certainly be. Binge-Eating Disorder will certainly be as an official category. There’s interest in compulsive sexual behavior. There’s even some interest in compulsive buying. I think those will be troublesome because people will say, ‘You know, you’re just expanding it to anything that’s considered excessive.’ There’s even Internet Addiction. Well, I mean people have talked about it. They’ve written about it. I don’t think it’s going to happen, but …

Lane: What’s your sense, then, of viable or acceptable cut-off points to these proposals?

Spitzer: I think it’s partly a question of how disabling … Is there really a condition? Are there people who really are addicted to Internet stuff analogous to other kinds of addiction. There probably are. At what point you put the threshold, I don’t know. Also, it’s a function of ‘Do you have a treatment?’ If you have a treatment, you’re more interested in getting the category in. If you have no treatment for it, there’s not as much pressure to put the thing in …

You’re asking, ‘At what point do you put a new category that seems to be on a continuum and looks like it’s an exaggerated form of something that’s normal’? I don’t know the answer to that.

But you know we had the same problem with gambling. Gambling is a little easier, because it’s become so obviously impairing and disabling. I don’t know what will happen with computer or sexual addiction.

Lane: Is there ever a point at which these various subsets of a disorder collapse back into their central feature—say, obsessive-compulsive behavior? We don’t need to define the various forms it will take, in short, because that’s already included …

Spitzer: Well, I think it’s different than OCD. It’s compulsive but doesn’t have the ego-alien quality. I mean, the person with OCD doesn’t enjoy washing their hands, because they’re afraid that if they don’t wash their hands they’ll come down with a disease. That’s very different from the person who wants to do any of these other things: gamble or … [trails off].

Lane: Yes, the pleasure-unpleasure dynamic may be harder to establish there … I did read that in NeuroPsychiatry Reviews there was a discussion about whether apathy should be included …

Spitzer: Well, why? Whether there’s an apathy disorder?

Lane: Yes, one largely due to the side effects of SSRIs. This seems to me tautological. If the drugs are creating side effects, that shouldn’t be grounds for inclusion as a new disorder.

Spitzer: Yes, it’s not a disorder, it’s a side effect.

Lane: It’s also a medical problem. There are clear medical reasons why the cholinergic and serotonergic systems would be strongly affected, especially following rapid withdrawal … I get alarmed when I hear side effects …

Spitzer: … mistaken for a disorder? I would … I would agree with that. I think there are some people who think there are primary apathy disorders. Maybe some neurological something, I don’t know.

The one thing I know about DSM-5 is that I will have nothing to do with it. Which is something I look forward to. [Laughs.]

Lane: Thank you for your time.


Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.


Mad in America has made some changes to the commenting process. You no longer need to login or create an account on our site to comment. The only information needed is your name, email and comment text. Comments made with an account prior to this change will remain visible on the site.


  1. I couldn’t get into this article and, frankly, I didn’t read most of it. Spitzer seems to accept that the DSM is a valid way to categorize psychiatric problems. I don’t.

    At one point he states, “this thing that we were influenced by pharmaceuticals is something that I just, I say, is just absurd.” I think all the DSMs reek of pharmaceutical influence.

    In addition, even though I like reading different points of view, I found this interview boring.

    Report comment

  2. Impressive interview, rich in detail about consequential changes in American psychiatry that are often obscured, not least by the central players themselves. When pressed, Spitzer has such flimsy justification for adding what became major new disorders affecting millions. The absence of any credible scientific basis for DSM inclusion is on full display here — it’s just about supporting his colleagues’ pet projects and hare-brained schemes. Frightening and telling. Well done for exposing.

    Report comment

  3. J. Sunda, well said, I agree. From a historical perspective, this interview is fascinating and illuminating. From a scientific perspective, it is horrifying. Spitzer and his DSM-III team basically just invented diagnoses and criteria sets based on a combination of politics, tradition, and personal bias. Science didn’t seem to enter the picture and to the extent it did, its findings were filtered through political negotiations and task force members pet ideas, rendering it moot. What is shocking is that these diagnoses and criteria sets remain all but unchanged 40 years after the were published. Major depressive disorder is still diagnosed by having 5 out of the 9 symptoms for at least 2 weeks that were invented out of thin air by Spitzer and his team. Except now, major depressive is a “real medical condition” and diagnosing and treating it is “evidence-based.” It is a triumph of marketing and perverse incentives over science and reality.

    Report comment

  4. Agreed with J. Sunada. I enjoyed the topics brought up by Lane and found his questions to be compelling, although at times Spitzer seemed to brush them aside or did not appear to realize the importance of the subject unless it struck a particular nerve with him. Some of the responses Spitzer gave left me a bit dumbfounded as well; that he places so little value in discussing whether or not the term ‘reaction’ has meaning is surprising to me among other terms or words given so little regard. Perhaps it shouldn’t surprise me. The justifications for new ‘disorders’ are indeed flimsy, it is disheartening to read when I can remember back to being a young patient and being told that psychiatry was based on strong evidence-based research.
    I did laugh at Spitzer acknowledging that social phobia and avoidant personality disorder are not different – they were included to make certain people happy. How many other disorders were added for similar reasons? Truly I found the whole interview worth reading just to read these words from one of the authors. Everything from the talk of miss-match between results in a clinical setting vs the real world, the theoretical tug of war between the different psych groups, the apparent mutual agreement that medication side effects should not be mistaken for a disorder (which we all know is happening), to Spitzer being apparently quite fine with his lack of involvement in the 5th edition. Thanks for sharing this Lane, I’m glad this interview was found.

    Report comment

  5. Absolutely chilling. The flippant way he throws around theories and pet topics, and then creates diagnoses out of them that will profoundly affect millions of people is just shocking. Complete disregard for the effect on the patients who will have to suffer the consequences.

    Imagine a surgeon using the same approach to conduct surgery, or an engineer trying out various theories on different constructions, or an airline pilot playing with the controls (“Let’s see what happens if I push this button!”).

    Sickening and shocking.

    Report comment

  6. Christopher,

    Thanks for the historical perspective. The section on dropping the word “reaction” was particularly interesting. Spitzer just didn’t seem to get it.

    But the move was definitely noticed in other circles. Here’s a quote from Ullmann and Krasner’s classic textbook: A Psychological Approach to Abnormal Behavior: Second Edition, 1975:

    “There are some features of DSM-II that seem to point in a more conservative direction. A major one is the change from a concept of a reaction (behavior) that is psychotic or neurotic to calling the behavior psychosis or neurosis with the implication of a disease entity.” (p 27)

    Report comment

    • Thanks for your comment, Philip. Yes, Spitzer completely minimized what was a radical move—refused to acknowledge its major implications, then as now. I didn’t know the Ullmann and Krasner quote but agree with it and appreciate it. Will use it more myself, with full credit to you! Thanks also for your very helpful critiques of DSM over the years, which also have shaped mine. Big thanks to you.

      Report comment

  7. Thanks Phil, you make a very important point. By dropping the word “reaction,” apparently without thinking much about the implications of doing so, Spitzer and his DSM-III effectively turned psychological effects into causes. Anxiety for example, when construed as a reaction, is understand as an effect of causes. This naturally leads the clinician to explore those causes and regard them as the real problem, with anxiety essentially a side effect.

    With a stroke of his pen, Spitzer turned anxiety and other reactions from effects into causes. People were now anxious because they had an anxiety disorder. Clinicians became free to disregard the context in which anxiety occurred, assume it must be the product of an illness in the individual, and employ interventions to reduce or eliminate anxiety itself without regard to its cause(s). The seemingly innocuous linguistic change of removing the word “reaction” paved the way for psychiatry to detach itself from psychoanalysis, promote disease theories of its new diagnoses, and go all-in on using psychoactive drugs with the goal of “symptom reduction.”

    Report comment

    • “With a stroke of his pen, Spitzer turned anxiety and other reactions from effects into causes. People were now anxious because they had an anxiety disorder.” That is spot-on, Brett, and exactly right—it was (and remains) a major conceptual shift in how we think about and diagnose anxiety. Unfortunately, it’s no surprise to me that Spitzer downplayed the enormity of the shift, to make it seem as if a “reaction TO” something were no longer meaningful. (‘Dropping the word reaction doesn’t really mean anything.’ I think that’s probably true.’) Remarkable in itself. Thanks for your great comments.

      Report comment

  8. Spitzer asks me at one point “if I’m familiar with the work of Herb Kutchins and Stuart Kirk.” Why yes, their books on DSM (The Selling of DSM [1992] and Making Us Crazy [1997]) are classics in sociology—outstanding contributions to the scholarship—and helped to drive my own investigation into the DSM archives. I quote them several times in my book Shyness. My favorite line from Making Us Crazy …?

    “By simply altering slightly the wording of a criterion, the duration for which a symptom must be experienced in order to satisfy a criterion, or the number of criteria used to establish a diagnosis, the prevalence rates in the US will rise and fall as erratically as the stock market” (1997, 244).

    Please keep that front-and-center as you read Spitzer’s account of how he and colleagues separated out the 7 anxiety disorders included for the first time in DSM-III. A fuller, more-detailed account of what happened appears in my book.

    Report comment

  9. Thanks for all comments analyzing this interview.

    I just finished reading Gary Greensberg’s Book of Woe about the DSM, not the first critique of the “bible” of psychiatry I’ve read. I must boast: no one needed to tell me as an adult that the DSM is nonsense, but I have sought professional critiques.

    As a victim of criminal psychiatry, violations of the state mental health code (violations in addition to nonsense psychiatric moves), I can’t get over my bad luck. I was violated at a religious hospital, built by the nuns who raised me and of course believed in the Christian bible. Meanwhile, seems pretty clear to me that if Jesus were born now, he’d be locked up (not crucified) based on the DSM. Seems the faithfuls’ deity is the first historical schizophrenic with god delusions, to this non believer, bible to bible.

    Meanwhile, a doctor who never met me, never evaluated me in the emergency room, labeled me suicidal and bipolar, a condition that does not exist, based on nonsense bias and counting games, based on non-evaluation, and transferred me to the mental ward involuntarily unconscious. I had said I was trying to “save my life” that day on Facebook. I had been set up by my EEOC documented psychologically hostile employer, but my true story was disbelieved by a first year intern who was not supervised.

    And there has been zero protection for me. Only fallout which continues nine years later.

    Report comment