Skip to Main Content

The Diagnostic and Statistical Manual of Mental Disorders — also known as the DSM and sometimes erroneously called the bible of psychiatry — is the authoritative guide to diagnosing depression, schizophrenia, and over a hundred other mental disorders. The latest version, published in 2013, was 19 years in the making. A new revision process put in place by the American Psychiatric Association, the DSM’s publisher, aims to make it easier for the manual to reflect changes in psychiatry.

Health care professionals in the United States and around the world rely on the DSM for descriptions and operational definitions for diagnosing mental disorders. It provides a common language clinicians can use to communicate about their patients. In addition, researchers can use its reliable definitions to help develop new medications and other treatments for mental health problems. Each edition of the DSM has sold over 1 million copies, indicating that it is also used by students learning about mental disorders as well as patients and their families who want to gain a better understanding of mental illnesses.


Revising the entire DSM every so often — the manual has been updated seven times since it was first published in 1952 — ensures that it keeps abreast with scientific developments in psychiatry. The overarching goal of these updates is to improve the manual’s validity, reliability, and clinical usefulness. Creating an update is a long, labor-intensive affair involving committees of experts with specific assignments aimed at overhauling the entire classification system.

Revising the entire manual at once helped ensure the uniformity and stability of the diagnostic information while that edition of the manual was in effect. But it made it difficult to incorporate new scientific knowledge in a timely fashion.

Advances in digital publishing that allow for instantaneous dissemination of changes at minimal cost have paved the way toward adopting a continuous improvement model for the DSM. Rather than updating all sections simultaneously, single entries can be updated pegged to specific scientific advances. Rather than waiting until the next scheduled revision to add a clinically useful change, such as incorporating a well-validated biomarker into the definition of a disorder, a change could be put into effect as soon as its clinical value has been verified.


This continuous improvement model will be entirely data driven. That will discourage changes that are not well supported by empirical evidence.

As my colleagues Dr. Kenneth S. Kendler and Dr. Ellen Leibenluft and I wrote last month in JAMA Psychiatry, the American Psychiatric Association (APA) is setting up a new web portal to field proposals for changes to the DSM. Individuals or groups making proposals must provide supportive information such as reasons for the change, data documenting improvements in validity, evidence of reliability and clinical utility, and a discussion of potentially harmful consequences associated with the proposed change. Proposals must also include a thorough review of the relevant literature and any secondary data analyses the proposers have conducted.

A steering committee appointed by the APA has defined four types of proposals: revising an existing diagnostic criteria set, adding a new diagnostic category, deleting an existing category, and correcting errors or inconsistencies.

We anticipate that most submissions will come from interested individuals, such as psychiatric researchers and clinicians, or from organizations such as psychiatric subspecialty groups and advocacy organizations. All of these submitters are separate from the committee overseeing the revision process. This is in marked contrast to earlier DSM revisions, in which the proposals were drafted by members of workgroups assembled by the APA, who were also responsible for reviewing the literature and reanalyzing data.

Putting in place a rigorous continuous improvement process for the DSM will let the manual reflect scientific advances more rapidly than was possible with the old revision process. This should ultimately result in a more valid and clinically useful diagnostic classification, which will help mental health providers and other clinicians to more accurately diagnose and treat their patients.

Michael B. First, MD, is professor of clinical psychiatry at Columbia University and a member of the American Psychiatry Association’s DSM Steering Committee.

  • The DSM describes “anti-social” behaviors that express emotional suffering and “socially unacceptable” reactions to the suffering and tags them as pathological without any scientific support. The APA established a task force in 2000 with a mandate to establish a scientific foundation for the DSM-5. Instead of admitting their failure, the new DSM changed the definition of a “mental disorder” for increased self-promotion and increased obfuscating. Few psychologists or psychiatrists noticed the change in definition of a “mental disorder;” this is the antithesis of science. I contend that depression naturally causes nervous tissue to atrophy from reduced usage (similar to all other body tissue) and thereby causes enlarged ventricles.

  • Dr. First – (this is in response to your reply about “enlarged ventricles” being an example of structural abnormalities found in people diagnosed w/ schizophrenia. I thought this was long ago refuted in that the studies failed to distinguish between the effects of purported disease process and the iatrogenic impacts of neuroleptic exposure. Nancy Andreasen’s studies conclusively point to the reductions in brain volume due to use of anti-psychotics. Research attempting to correlate brain abnormalities & notoriously unreliable diagnoses has largely been spurious.

  • I recently came across an academic article that talks about the illusory theoretical underpinnings of psychiatry – it describes how current psychiatric trainees and junior psychiatrists blindly accept the methods used in psychiatry because they simply (wrongly) assume that much smarter people before them have somehow sorted out all the details relating to the theories underlying psychiatry. The article also says that the teachers who instruct psychiatry are not lying (i.e., are not “bad people”), but are merely repeating what they have learnt without questioning.
    Here’s the article: McLaren, N. (2016). Psychiatry as Bullsh*t. Ethical Human Psychology and Psychiatry, 18(1), 48-57.

  • The DSM is a joke. A book of guesses, based on cultural bias, societal norms and even personal opinons that change as often as the political environment and society values do. Psychiatry’s continued insistence to use it as their bible makes the profession even more laughable than it already is. There is no existing scientific evidence of what actually causes mental illness, what treats mental illness or what cures mental illness. Anti-depressants and other psychotropic drugs are a scheme created by big pharma solely to line their pockets with more money. There is not one shred of proven science that they even work or why. Approximately 40% of patients don’t respond to them. Shame on every one of you who are so eager to chemically alter the brains of your patients with little thought to unknown consequences. A day will come when this practice is looked upon as the horror it is. Now they want to make the DSM easier to add new “diagnosis”, leading to pharma coming up with new “cures”. The medical establishment has long needed to separate themselves from the practice of psychiatry because it implies legitimacy where there really is none.

  • Dr. First:
    I do have a question: When will officers at the DSM-5 decide to investigate what William Bernet has been saying and doing about PAS and the DSM-5?

    After reading your piece about new rules / regulations, I am very concerned that William Bernet and his associates will explore novel ways to force the inclusion of his version of PAS (PAD / PA) in the DSM-5. Ironically, after the publication of the DSM-5 in 2013, Bernet has affirmed, authoritatively, in a variety of writing the scientificity of PAS because it is already included in the DSM-5. In Bernet’s own words:

    “The actual words of “parental alienation” are not in DSM-5, but there are several diagnoses that can be used in these cases. I would say the “spirit” of parental alienation is in DSM-5, even if the words are not.”

    Bernet and his followers have been repeating this mantra since 2013, when he was shocked that the DSM-5 had not included his version of PAS, which he calls ‘relational’.

    Despite the setback, he has continued to develop his organization of PAS as an instrument for proselytizing worldwide. Two of his most recent actions have explicitly aimed at strengthening the scientific and organizational features of PAS. The first involves a reorganization of his private foundation by extending the number of new international members (pasg). The link to his foundation is
    The second, and I believe most ominous, action involves the publication of his bibliography on PAS on the website of the Medical School and Vanderbilt, where he used to teach and where, for reasons unknown to officers and faculty that I have contacted, he has an email account which he uses as his professional account.

    Bernet had presented a previous version of the bibliography in his proposal for the DSM-5 that was rejected. The new bibliography, to which several his international followers have contributed, is now available as a ‘scientific bibliography’. In his own words written in the website:
    Parental Alienation Database

    “A large bibliography regarding parental alienation has been jointly developed by PASG and the Eskind Biomedical Library of Vanderbilt University Medical Center. Access this Parental Alienation Database at
    Thus, the Medical Center at Vanderbilt now legitimizes the scientificity of his PAS bibliography.

    Lastly, I find even more ominous the recent publication of an article in which Bernet & Co. present a new version of PAS and the DSM-5, which he calls CAPRD:

    A sentence in the abstract states: “A new condition, “child affected by parental relationship distress” (CAPRD), was introduced in the DSM-5. A relational problem, CAPRD is defined in the chapter of the DSM-5 under “Other Conditions That May Be a Focus of Clinical Attention.” The purpose of this article is to explain the usefulness of this new terminology.

    Now, I have been researching–pro bono and as volunteer–this PAS saga, and I find the entire story from 1985 to today, and every day I am dumbstruck by new developments in the saga linking PAS e DSM-5.

    Know that I’m not a psychiatrist (or psychologist). I am a political scientist who a year ago came across, by accident, the “phenomenon of PAS” in Italy and more broadly in Europe, the US, Latin America, and Australia.

    The fortuitous accident? Investigating problems concerning violence against women in the Italian setting, I began to read about the link between Richard Gardner and Columbia University. Columbia is my alma mater — where I have studied and taught. I honestly could not bring myself to believe what the web, articles and books were saying about Richard Gardner — stuff such as Gardner being the greatest psychiatrist that the Department of Psychiatry of Columbia, the best in the world, had ever produced, and under whose leadership had achieved levels of unimaginable sophistication, etc.

    Thus, my initial interest was political — how Gardner’s PAS had taken roots in Italy (an event that remains a mystery), and, thereafter the introduction of Bernet’s relational ‘theory’ and the creation of alliances in civil society and the state linking professional orders of psychologists / psychiatrist, of lawyers and Courts, of politicians, and of social workers. This alliance now has nearly total control of the judicial-legal process involving PAS cases–which in Italy are a humongous 35,000. That is, at least 35,000 children (or 35,000 families) have been involved in Court proceedings disputing PAS cases in the name of the DSM-5. Of those 35,000 cases, nearly 95% see the father / husband as alienated and hence accusing the mother/wife as alienating. More disturbing — at least for me when I read published PAS materials and court sentences — is that a great majority of cases (~ 75%) involve accusations of sexual abuse of children against the ‘alienated father’. And even more disturbing are the results of recent empirical research on actual PAS cases, which show that nearly 92% of cases involving accusations of sexual abuse are legally founded. That is the accused is found guilty. Mind-blowing!

    Personally, I have offered my help and support as a social scientist who is strongly rooted in logic and methodology, including statistical methods. (I taught statistical methods for three years in the Statistics Department at Columbia.) My knowledge of English as well as of the US academic system, about which most people are ignorant, has helped in organizing a movement of individuals and organizations — mostly women — which have interests in problems of violence against women … many women see the PAS-DSM5 as instrument contributing to violence against women (and children). I have also directed my investigation to the ‘other side’: the organizations of fathers / men and their significant contributions — organizational and financial — which support PAS and the presumed scientific links to the DSM-5.

    My initial query stand: what can the directorate of the DSM-5 to confront Bernet’s manipulation e abuse of the DSM-5?

    I now add: should the directorate of the DSM-5 be concerned with the potential abuse of the new rules and regulations by types like Bernet?

    • Salvatore,

      You are making sound scientific arguments about an unscientific subject. Every diagnosis in the DSM is political; new entries (or discarded ones) are only more obvious. The DSM causes tremendous social harm by pathologizing natural emotional suffering; you are witnessing the process with PAS.

      Best wishes, Steve Spiegel

  • We need to remember that despite decades of research, scientists have not been able to find any structural or other brain differences between people with and without DSM conditions (this is the reason why these ‘disorders’ are diagnosed using checklists – there are no objective tests). Also, when people are given a fancy psychiatric label (from the DSM) and are also told “these are long-term conditions” – that itself can trigger nocebo effects (which are the opposite of placebo effects), further complicating the condition of the patient. I think we really need to stop giving labels and medicines (that pharmaceutical companies come up with using a hit or miss approach). Additionally, psychiatric drugs do more harm than good – I came across the following report titled “The Case Against Antipsychotics: A Review of Their Long-term Effects” (it is accessible online).
    Mental health issues originate as a result of human experience and these experiences (such as various psychological stresses, along with the resultant mental proliferations and rumination that can follow) can be effectively treated with various psychological interventions.

    • While it is true that when diagnosing an individual patient, imaging and other laboratory tests and not specific enough to help in making a diagnosis, there are years of research showing associations between the diagnostic entities in the DSM and structural and functional abnormalities in the brains of groups of people with a particular disorder. For example, if you look at average ventricular size on structural brain MRI’s of 1000 patients with a DSM diagnosis of Schizophrenia and compare them to the average ventricular size of 1000 patients without any diagnosis, on average the ventricular size of the patients with a diagnosis of Schizophrenia are enlarged. (The ventricle is the normal space in the center of the brain that is filled with fluid that is visible on MRI). This establishes that the DSM diagnosis of Schizophrenia is a real thing in nature and not something that has been made up out of whole cloth by the American Psychiatric Association. What you cannot do is look at an individual’s patient MRI results and definitively make a diagnosis of Schizophrenia because there are patients with Schizophrenia who have normal ventricle sizes and patients without any diagnosis who have enlarged ventricles.

    • Hi Michael First: Since there was no ‘reply’ button for your post, I am replying to my own post.
      When someone is subjected to a great deal of psychological stress, the structure of their brain changes. Numerous studies including animal studies have shown this (see for example the following two articles published in Nature: and article titled “Social influences on neuroplasticity: stress and interventions to promote well-being” published in Nature Neuroscience, 15, 689–695). But these structural changes are reversible (neuroplasticity) through psychological means. They are reversible even in mice (see the above references).
      As I mentioned earlier, mental health issues originate as a result of human experience and these experiences (various psychological stresses, along with the resultant mental proliferations and rumination that can follow) bring about gradual changes in the structure of the brain. These changes happen through neuroplastic and epigenetic mechanisms. To reverse these changes, various psychological interventions help. For example, many studies have shown that the practice of mindfulness (that prevents mental proliferations and rumination and reduce psychological stress) changes the structure and function of the brain in positive ways.

    • Michael First: There are plenty of associations between one psychiatric “condition” and another set of physiological conditions. These associations are very close to meaningless in terms of defining actual diseases/disorders, because the fact that, for instance, it is more LIKELY that a certain gene is present in a certain diagnosed person generally means that many people with the condition don’t have the gene and many people with the gene don’t have the condition. It also ignores the question of causality: do more depressed people have a certain brain scan because their brain is malfunctioning, or does having depressed thoughts lead to those changes in the brain? The complete lack of consideration of causality in the DSM makes the DSM categories worse than useless – they are misleading, preventing actual research on any subset of depressed (or whatever) people who MIGHT actually have something physiologically wrong.

      Essentially, it is a fool’s errand to define psychiatric disorders based on symptoms alone. This is what the DSM does, and the result is an ever-expanding list of claimed “disorders” that have less and less connection to any kind of physiological reality at all. To make matters worse, the inevitably subjective boundaries between “normal” and “disordered” allows for the “disordered” line to creep slowly but inexorably closer and closer to “normal,” such that now almost everyone can qualify for some diagnosis or other. This works great if you’re a pharmaceutical company trying to make money, but from a scientific viewpoint, the DSM is an unqualified disaster, and needs to be scrapped. Even the head of the NIMH admitted the same.

  • Now pharma can immediately pay-off the APA to name a niche behavior as a mental disorder which can be immediately branded with a particular psychotropic drug. Instead of all the time it took to create Social Anxiety Disorder and brand it with FDA consent for Paxil exclusively, they can do this immediately without having to got through the time and scrutiny of creating a new manual. This leads to more dangerous fraud.

  • What was “scientific” about declaring homosexuality was not a disorder when previously psychiatry saw it as such? When you are you going to stop playing the scientism game and admit that a lot of psychiatry is culturally bound?

    • Homosexuality was removed from the DSM in 1972 at a time when the DSM was admittedly not very scientific. The homosexuality story, while important, is not relevant to the way the DSM has been revised since the 1980’s.

Comments are closed.