Nassir Ghaemi M.D., M.P.H.

Mood Swings

The secret of DSM-IV: The danger for DSM-5

How DSM-IV attitudes could harm DSM 5

Posted Apr 28, 2010

John Kenneth Galbraith once said that all academic disciplines have their disputes, much cherished by participants, and regularly combining differences in method with deep personal dislike.  I continue this debate reluctantly, because I have no personal acquaintance of Dr. Frances, nor he of me, and the internet always magnifies emotional tone.  Nonetheless, I will continue the discussion because of its practical importance for psychiatry and the public.

When DSM-IV was published in 1994, I was just graduating my psychiatry residency; I have lived the first half of my career under its hegemony.  In 1994, Allen Frances was the chairman of the entire DSM-IV task force and head of the department of psychiatry at Duke.  This is, to some extent, a debate, about the future, between the past and present of psychiatry. (If experts are biased and need to be taken with grains of salt, as my colleague states, we should keep in mind that he is an expert in diagnostic systems; I was not asked to be part of the DSM-5 task force). 

There is an old saying that you should be nice to others on your way up because you will see them on your way down.  With DSM-IV, at the pinnacle of his power, my colleague was not so nice to people with bipolar disorder.  He did allow for hypomania to be diagnosed, this is true, but at the cost of multiple other restrictions to avoid the boogeyman of overdiagnosis of bipolar disorder. The restrictions were as follows: the 4 day minimum criterion for hypomania, a narrowing of the definition of mixed episodes (and a corresponding broadening of MDD and bipolar depression), and the exclusion of antidepressant-induced mania as defining bipolar disorder (which is another broadening of MDD).  (I have provided a literal example of death as a result these DSM-IV falsehoods).  In the last 16 years, the scientific evidence has accumulated notably against all of these restrictions.  But my colleague waves his hand impatiently at science and experts.  Now, later in his career, meeting again those he saw on the way up, he remains unfriendly to persons with bipolar disorder.  This is unfortunate, not for him, but for those who are harmed by the preference for personal opinion over scientific knowledge.

It was not supposed to be this way.  Somewhere along the line from DSM-III to DSM-IV, the leaders of psychiatry lost their way. In 1980, Robert Spitzer and Gerald Klerman and others who led the major revolution of DSM-III promised that science would have top priority in all revisions in the future. (By the way, contrary to Frances' claims, the schizophrenia experts in DSM-III advocated narrowing that diagnosis; I, as a mood disorders and psychopharmacology researcher, have long advocated narrowing the MDD diagnosis and using fewer antidepressants and antipsychotics). The fact that one had to pay attention to "pragmatic" (which Dr. Frances prefers to the term "political") considerations - the beliefs of clinicians, the wishes of patients, our general ignorance about many scientific facts, the limitations of our treatments, the needs of insurance reimbursement - was accepted as a necessary evil.  

I have learned something new and disconcerting from the leader of DSM-IV: He sees this necessary evil as a virtue.  Here is the secret to DSM-IV, a major reason for our problems today; its founder is now pellucid about it: Science is, and should be, the least important factor in the DSM system;  pragmatic, professional, economic, social (dare one add it all up, as Aristotle did, in the word "political") considerations are more important.

Science matters least. 

This is an excellent lesson to teach all of our psychiatric residents now, so that they do not enter our field with our past illusions.  We actually thought our leaders wanted to know the truth. How silly of us.

Now that I know the truth does not matter (perhaps my colleague has imbibed our cultural postmodernist disease too)  I realize now that this is not a scientific discussion, primarily, it is a philosophical one.  So let me try to reconstruct the logic of Dr. Frances' argument premise by premise, with my comments:

Premise 1. Antipsychotics are overused and harmful.

This may be true, but did my colleague ever ask himself why antipsychotics are overused? And does he not think that antidepressants are overused and harmful?  The study I previously cited showed that, in psychiatric practices, over 60% of patients receive antidepressants, only about 25% receive antipsychotics. Which seems overused?  (And, by the way, this is in thousands of private practices across the country, not the rarefied confines of ivory tower researchers; the same holds for the studies of hypomania and bipolar underdiagnosis I had cited).

He blames the pharmaceutical industry, which manipulates doctors and patients. (An easy target these days: Agreed.) He also implies that doctors are just poor practitioners and not easily taught to practice well. (A bit dismissive of our colleagues, it seems to me, though not entirely untrue.) These are not the only causes; another cause is the DSM system as currently bequeathed to us by my colleague (more on this below).

Premise 2. The science behind the data on hypomania is too weak to be definitive enough for a diagnostic change.

This is simply false, as an empirical scientific matter, as I stated in my previous blog post. As he is apparently happy to avoid the label, my colleague admits he is not an expert on bipolar disorder, so perhaps he should review that evidence more closely.

Premise 3. When the science is weak (which in psychiatry means most of the time), DSM should err on the side of narrow definitions to avoid overdiagnosis and overtreatment.

This seems quite reasonable, and agrees with the current zeitgeist in American society. But I wonder why my colleague expresses no concern about the quite broad definition of major depressive disorder (MDD) and consequent overuse of antidepressants? In fact, this whole debate about bipolar disorder type II is not about diagnosing new patients with a mental illness who are not currently diagnosed with any mental illness. These patients are either going to be diagnosed with MDD or bipolar disorder type II. They all have major depressive episodes; the only question is whether they also have hypomanic episodes.
Today, DSM-IV, for which our colleague was the responsible leader,  diagnoses MDD broadly, and bipolar disorder type II narrowly. I suggest broadening the bipolar type II definition and correspondingly narrowing the MDD definition. It is not a matter of being generally broad or generally narrow; my colleague is narrow for some conditions, but quite broad for others. Why be broad in one case and narrow in another? Further, he takes this approach despite the preponderance of the scientific evidence indicating otherwise.He might respond that as a practical matter, overuse of antidepressants is less risky than overuse of antipsychotics. This is somewhat questionable, since about half of our new antipsychotics have little or no cardiac or obesity risks, while the antidepressants seem to cause suicide in some persons - besides other important risks, such as causing mania in those with type II bipolar disorder, or simply being ineffective for depression in those persons. (Again one could cite numerous scientific studies to support these facts, but science matters least). 

Conclusion. DSM-5 bipolar disorder type II should not be broadened. 

One can see why he and I come to different conclusions.

To summarize: It is now clear how DSM-IV was created in such a manner that psychiatry has been doomed from the start to being an unscientific and anarchic profession.

Let's state a fact: Most clinicians and patients do not trust DSM. They see our psychiatric nosology as a made-up system, meant to enrich doctors and drug companies and hospitals, not a science of diagnosis that is getting us closer to better definition of disease. Dr. Frances is agreeing with this view, and taking it a step further. Not only is DSM mainly a pragmatic (social/economic/political) entity, it should be so.  This is quite a claim. Even in the presence of sufficient scientific evidence otherwise, DSM definitions should primarily aim at controlling the mistaken practice of ignorant doctors and patients.  

But the cause is the consequence: Clinicians do not trust the diagnoses; so they prescribe drugs for symptoms, resulting in overuse of medications. Because of "pragmatic" concerns, DSM seems to be knowingly and willfully wrong in its definitions. No wonder we get nowhere in our biological and etiological research.

 Many clinicians and the public, fed up with the errors and falsehoods of DSM-IV, will have the urge to feel supportive of Dr. Frances' opposition to changes in DSM-5.   They remind me of those who oppose health care reform on the grounds that our system is so broken, any further change must be for the worse. They do not seem to realize that the current mess was caused by those who now so loudly denounce any attempt to fix it. 

About the Author

Nassir Ghaemi, M.D., M.P.H.,

is Professor of Psychiatry at Tufts University School of Medicine, and Director of the Mood Disorders Program at Tufts Medical Center in Boston.

More Posts