Opening Pandora’s Box: The 19 Worst Suggestions For DSM5

By Allen Frances, MD | den 11 februari 2010

Dr Frances was the chair of the DSM-IV Task Force and of the department of psychiatry at Duke University School of Medicine, Durham, NC. He is currently professor emeritus at Duke.

I have previously criticized the DSM5 process―for its unnecessary secretiveness, its risky ambitions, its disorganized methods, and its unrealistic deadlines.1-6 Now, it is finally time to evaluate the first draft of the recently posted DSM5 product (at

Poor and inconsistent writing

Perhaps it should occasion no surprise that a flawed process should yield a flawed product. The most fundamental problem is the poor and inconsistent writing. Admittedly, early Work Group drafts are often written imprecisely and with varying quality, but it is surprising that the DSM5 leadership has failed to edit for clarity and consistency. It would be a waste of effort, time, and money to conduct field trials before the new criteria sets receive extensive revision. The poor writing is also a bad prognostic sign, suggesting that the DSM5 text sections for the various disorders may eventually be equally inconsistent, variable in quality, and sometimes incoherent.

Higher rates of mental disorder

In terms of content, most concerning are the many suggestions for DSM5 that would dramatically raise the rates of mental disorder. These come in 2 forms:

1. New diagnoses that would be extremely common in the general population (especially after marketing by an ever alert pharmaceutical industry)

2. Lowered diagnostic thresholds for many of the existing disorders.

DSM5 would create tens of millions of newly misidentified false positive “patients,” thus greatly exacerbating the problems caused already by an overly inclusive DSM4.7 There would be massive overtreatment with medications that are unnecessary, expensive, and often quite harmful. DSM5 appears to be promoting what we have most feared--the inclusion of many normal variants under the rubric of mental illness, with the result that the core concept of "mental disorder" is greatly undermined.

Unforeseen consequences

A third pervasive weakness in the DSM5 options is their insensitivity to possible misuse in forensic settings. Work Group members cannot be expected to anticipate the many ways lawyers will try to twist their good intentions, but it is incumbent on the DSM5 leadership to establish a thorough ongoing forensic review that would identify the many likely instances of proposals with important forensic implications (for example, the expansion of pedophilia to include attraction to adolescents).

Space constraints (as well as my own blind spots and limitations in expertise) make this a limited survey, both in the numbers of issues discussed and the depth of discussion possible on each. I would encourage the field to identify the additional problems that will require correction.


The Psychosis Risk Syndrome is certainly the most worrisome of all the suggestions made for DSM5. The false positive rate would be alarming―70% to 75% in the most careful studies and likely to be much higher once the diagnosis is official, in general use, and becomes a target for drug companies.8 Hundreds of thousands of teenagers and young adults (especially, it turns out, those on Medicaid) would receive the unnecessary prescription of atypical antipsychotic drugs.9 There is no proof that the atypical antipsychotics prevent psychotic episodes, but they do most certainly cause large and rapid weight gains (see the recent FDA warning) and are associated with reduced life expectancy―to say nothing about their high cost, other side effects, and stigma.

This suggestion could lead to a public health catastrophe and no field trial could possibly justify its inclusion as an official diagnosis. The attempt at early identification and treatment of at risk individuals is well meaning, but dangerously premature. We must wait until there is a specific diagnostic test and a safe treatment.

Mixed Anxiety Depressive Disorder taps nonspecific symptoms that are widely distributed in the general population and would therefore immediately become one of the most common of all the mental disorders in DSM5. Naturally, its rapid rise to epidemic proportions would be ably assisted by pharmaceutical marketing. It is likely that medication would not be much more effective than placebo because of the high placebo response rates in milder disorders.10

Minor Neurocognitive Disorder is defined by nonspecific symptoms of reduced cognitive performance that are very common (perhaps almost ubiquitous) in people over fifty. To protect against false positives, there is a criterion that requires objective cognitive assessment to confirm that the individual has decreased cognitive performance, but getting a meaningful reference point is impossible in most instances and the threshold has been set to include a whopping 13.5% of the population (ie, the percent of population within the first and second standard deviation). Moreover, the suggestion for objective testing will probably be widely ignored in the primary care settings where the bulk of diagnosing will be done.

Medicalizing the expectable cognitive impairments of aging will result in much unnecessary treatment with ineffective prescription drugs and quack folk remedies. These will undoubtedly attain great popularity since there will likely be a very high placebo response rate.

Binge Eating Disorder will have a rate in the general population (estimated at 6%) and this will probably become much higher when the diagnosis becomes popular and is made in primary care settings. The tens of millions of people who binge eat once a week for 3 months would suddenly have a “mental disorder”― subjecting them to stigma and medications with unproven efficacy.

Temper Dysfunctional Disorder with Dysphoria is one of the most dangerous and poorly conceived suggestions for DSM5―a misguided medicalization of temper outbursts. The “diagnosis” would be very common at every age in the general population and would promote a large expansion in the use of antipsychotic medications, with all of the serious attendant risks described above. Apparently, the Work Group was trying to correct excessive diagnosis of childhood bipolar disorder―but its suggestion is so poorly written that it could not possibly accomplish this goal and instead would it would create a new monster.

The misapplication of this diagnosis would provide a blanket excuse for reduced personal responsibility and will lead to forensic nightmares. It is a nonstarter.

Paraphilic Coercive Disorder would expand the pool of sex offenders who are eligible for indefinite civil commitment because they have a “mental disorder” to include cases of sexual coercion. Paraphilic Coercive Disorder was initially considered for inclusion in DSM-III-R (under the name Paraphilic Rapism) but was rejected because it was impossible to reliably and validly differentiate those rapists whose actions are the result of a paraphilia from the large majority of rapists who are motivated by other factors (such as power). Given the facts (acknowledged in the rationale section) that most rapists are savvy enough to deny sexual fantasies and the unreliability (and unavailability) of laboratory testing, the diagnosis will inevitably be based only on the person’s behavior, leading to a potentially alarming rate of false positives with consequent wrongful indefinite commitment.11

Hypersexuality Disorder would be a gift to false positive excuse seekers and potential forensic disaster. Another clear nonstarter.

A Behavioral Addictions category would be included with the substance addictions section and would start life with one disorder, Pathological Gambling (transferred from Impulse Disorders section). Next in line might be a new category for Internet Addiction. This could provide a slippery slope leading to the back door inclusion of a variety of silly and potentially harmful diagnoses (ie, “addictions” to shopping, sex, work, credit card debt, videogames etc, etc, etc) under the broad rubric of “behavioral addictions not otherwise specified.” The construct “Behavioral Addictions” represents a medicalization of life choices, provides a ready excuse for off loading personal responsibility, and would likely be misused in forensic settings.