The Dangers of DSM-5
The DSM-5 is the latest edition of the classification of all mental diseases published by the American Psychiatric Association. It includes nearly four hundred mental disorders. In the final stages of its development and until its publication in May 2013, the DSM-5 has been at the centre of a heated debate. Key among the many criticisms raised against it has been its alleged “danger” to public health.
I would like to examine this question of the danger of the DSM more closely. The fact is that it is not merely a squabble among experts, but a problem concerning the public at large.
Starting from its third version, which was published in the 1980s, the DSM, originally only an American classification to be used by statisticians and epidemiologists, has gradually become the reference tool for the entire field of global psychiatry: this includes research, especially pharmacological research, but also psychiatric research into the brain and brain imaging techniques; epidemiology; medico-economic evaluation; teaching; scientific literature; and especially — and increasingly so — clinical work. Its influence, though not total, is considerable, and it has de facto replaced the official classification of the World Health Organization, the so-called CIM 10, despite the latter being legally binding in almost all countries, including France.
Before discussing the question of the dangers inherent in the DSM-5, a few preliminary remarks on classifications and psychiatry are necessary.
No psychiatric classification can claim to be scientific because to this day, mental diseases have not been given any definite scientific explanation – neither causal nor psychopathological. There is no “biological marker” in psychiatry that we could rely on to establish a diagnosis. Diagnosis in psychiatry is an entirely clinical matter, which means that it is based on interviews and tests that can be more or less formalized, but definitely not on the complementary radiological or biological exams. The DSM method has replaced scientific proof with a consensus of experts: these are supposedly enlightened by studies published in the global scientific literature that comply with the criteria of pharmacological research (randomized clinical trials). This represents, together with other things, the model of evidence-based medicine.
Psychiatric classification is essential especially for epidemiological research; however, until the 1980s, a great disparity existed between different clinicians and countries. There was no common criteria, no common language to establish a diagnosis. Theoretical references, names of illnesses, and the clinical vocabulary varied considerably from one practitioner or country to another. This variation created insurmountable difficulties for research and epidemiology, which both require a common language and criteria in order to compare, calculate, carry out experiments, and evaluate the efficiency of a given molecule or help orientate public mental health policies. The DSM replaced this cacophony of the so-called classical clinical psychiatry with an “atheoretical” model, which comes with a list of superficial, easily observable types of behaviour and criteria of inclusion and exclusion. Like in a classical expert system, to obtain a psychiatric diagnosis we need only to follow a decision tree. “Scientific” language calls this “operational criteria.”
In psychiatry, it is impossible to establish an objective definition of normality – what is normal varies according to different contexts, cultures and societies. There is a tendency to make up for this “shortcoming” by resorting to the criteria of functionality: in other words, a behaviour or a symptom is considered “abnormal” if it creates enough difficulty for the individual. This difficulty is either felt by the subject himself or it manifests in his interaction with those that are close to him – e.g., his family – and in the larger social field, especially in professional life. The criterion of functionality is obviously very “fragile” and subjective, especially when we are dealing with children or adolescents, because it depends on the “tolerance” of others – the family, institutions and society in general – which is always variable. The DSM method tends to objectivize this criterion, as if a simple limit of tolerance could be equivalent to a biological threshold. Also, because the line that separates normal behaviour from abnormal is a matter of consensus, any slight change in this threshold becomes a medico-economic issue. Any new disorder can lead to the inclusion or exclusion of a large number of patients, which has economic repercussions such as for example the rise in prescriptions and benefits.
Any classification of mental disorders must satisfy at least three criteria – reliability, validity and usefulness. Reliability means that given the same patient, the same clinical case, two clinicians must have the best chance at establishing the same diagnosis. Validity requires that by using this classification and criteria a clinician can easily separate one disease or syndrome from another. And finally, the classification must be useful, in other words clinical practitioners, researchers and epidemiologists, as well as other actors in the field of mental health (insurance companies, healthcare administration and especially service users) must consider it a useful tool in their respective domains. The DSM-5 does not truly satisfy any of these three objectives – for details, I refer the reader to the well-documented article by François Gonon in L’Information Psychiatrique. (1)
Compared to other medical domains, psychiatric diagnosis has its own specificities. To take a banal example, announcing to someone that they have a mental disease is not and will never be the same as informing them that they have broken their arm. Of course there are some somatic diseases that must also be announced to the patient carefully and with sufficient empathy and tact, for example due to their prognosis, but that is another matter. In most cases, a psychiatric diagnosis has the effect of a stigma, especially in the case of children; a mental disease means that one is “unlike the others” or even “mad.” Conceivably, the belief that all mental diseases are caused by cerebral lesions might have the positive effect of reducing the risk of discrimination. If mental disease equals brain disease and the brain is an organ like any other, there is no longer any reason to discriminate against mental patients, or we would need to also discriminate against those suffering from cardiac disease or hepatitis… However, in reality, the process of stigmatisation does not solely derive from aetiology – what also counts is the social imaginary, the fear of madness, the strangeness of certain types of behaviour, and many other factors. This was the reason for the failure of a British public information campaign launched by the Royal College of Psychiatrists several years ago. Its theme was: “Do not be scared of mental disease – it’s a disease like any other.”(2) A psychiatric diagnosis can also give the mentally handicapped or their parents access to certain welfare benefits or services – an important factor at a time when users’ voices are gaining more ground, which has also had a number of positive effects.
Since the 1960s, there have been two main criticisms raised against psychiatry that have had an impact on its classificatory systems. Psychiatry was born in the historical context of modernity, which saw the triumph of reason over belief. One of the consequences was the incarceration of those deemed “unreasonable” by “the reasonable.” Hence the idea, developed by thinkers such as Roy Porter (3) and Michel Foucault (4), that psychiatry became a medical discipline only secondarily, while its initial vocation in the modern era was the confinement of the mad in asylums, where doctors played the role of wardens – a process Foucault calls “the great confinement.” This power to incarcerate, which is more or less recognized by the different national legislations, is an exception to, or a restriction of, individual freedom. Although involuntary commitment exists in all modern societies, some have compared forced treatment to torture, referring to a resolution passed by the UN Human Rights Council (5). Putting someone into a diagnostic category can therefore impact his or her personal liberties, which has led some to argue that all classification is bad and that in psychiatry we should completely refrain from diagnostics because to classify automatically means to confine, to open the first gate of the “asylum.” This is why psychiatric classifications have been fiercely criticised by anti-psychiatry.
The second reproach concerns psychiatric medication – psychotropic drugs. Because so far no specific aetiology of mental diseases such as depression or the psychoses has been found, the names “antidepressants” or “antipsychotics” seem a complete sham. Are these drugs effective in themselves or are we dealing with a placebo effect? This is an important question because their performance, at least in the case of antidepressants, has been shown to approximate that of a placebo. Their effect on the brain, especially on the neurotransmitters, is undeniable, but how should we understand it? Is it a form of cure that fixes a chemical imbalance caused by the illness, or does their calming or exciting effect simply create an artificial state, as any drug would do? Unless we assume an ideological stance, we can answer this question neither completely nor generally. Also, in cases of prolonged use, the long-term effects of these types of medication, in particular the latest class of antipsychotics (the so-called second-generation antipsychotics), have been subject to criticism and based on independent studies they have given rise to considerable reservations: they may include metabolic disease, diabetes, obesity, tardive dyskinesia, loss of brain tissue (6). Therefore, are psychotropic drugs useful, or are they in fact just as dangerous as the diseases they claim to treat? My answer is this: if correctly prescribed, they can be beneficial, in the opposite case and especially in children they can be worse than the disease or so-called disease they are supposedly fighting against.
Questions of causality have long been a controversy in psychiatry. Does mental disease have an organic, psychical or social origin? Prudence dictates that we keep all three factors as models that, depending on the context, can co-determine a mental illness to a different and varying degree. However, over the last thirty years, we have seen the rising hegemony of what is usually called the organic or biomedical model and a simultaneous moving away from the other two (psychical and social) models. The “brain” has become very fashionable, leading to a kind of “neuromania” and replacing the “brainless” psychiatry of psychoanalysis with the “soulless” psychiatry of neuroscience. Recently, we have nonetheless been witness to an opposite trend, which has to do with a certain disappointment with biological psychiatry (brain imaging, genetic research, etc.), which, contrary to expectations, has not uncovered any of the secrets of major mental diseases. No biological marker has been found and despite massive human and financial investment – which is set to continue given the recent “second decade of the brain” declared by president Obama – it has led to few useful applications for everyday psychiatric practice. Research should obviously continue. A number of scientific discoveries have indeed been made, but so far they have identified only correlations between the clinical picture of a disease and biochemical disturbances, rather than elements useful for diagnosis. Above all, research has not found any causal agents comparable to the bacteria or viruses that cause certain organic diseases.
Having reiterated some of these key points, which will no doubt sound familiar to those conversant with the issues of today’s psychiatry, let me now return to the potential dangers of DSM-5. The DSM-5 will definitely be responsible for three serious problems of public health, which are in fact mutually interdependent: overdiagnosis, overprescription, and overmedicalization. Overdiagnosis manifests itself mainly by an inflation of the number of mental diseases listed in the Manual. Over the last fifty years, this number has risen fourfold: from approximately a hundred listed in the DSM-2 to more than four hundred in the DSM-5. How can we explain this increase? We may think of a number of reasons.
Firstly, the idea that we are living in an increasingly pathogenic and stressful society, which is therefore likely to generate a growing number of mental disorders. I disagree with this argument, at least in relation to the Western society – not only because it is too general and inexact, but also because life in the societies of previous generations, fraught with wars and economic crises, carried no fewer risks for the mental health of the population. What is more likely to have changed is our relation to psychic suffering. What used to be considered a normal reaction to the accidents of life (separation, heartbreak, loss of employment, bereavement, etc.) is now seen as something requiring medical care and treatment. I will come back to this when I speak of over-medicalization.
Another way of explaining this inflation of mental disease is to posit that these “new” disorders had already existed — i.e., there has been no increase in their incidence — but we had previously not been able to identify and describe them. That is in part true: clinical and therapeutic trials or certain radiological data, etc., have indeed greatly facilitated the identification and definition of some disorders, as well as allowing us to better distinguish between different pathologies. However, this progress has had only little influence on the diagnostic inflation.
The third explanation, which I will side with and discuss in more detail, invokes the responsibility of the DSM. Starting from the 1960s and especially the 1980s, which saw the arrival of drugs with fewer side effects, psychotropic medication has had a growing impact on psychiatric practice. First of all in hospitals, where the drugs’ sedative effects have played a positive and liberating role, but also increasingly in private practice, where they have allowed clinicians, especially general practitioners, to offer a quick medical answer to “existential difficulties.”
At the same time, their efficacy has had a strong impact on the definition and grouping of symptoms. For example, in the case of anxiety this has lead to a nosographic distinction between types of anxiety that respond well to antidepressants (panic attacks) and those that respond to tranquilizers (generalized anxiety disorder). This new definition replaced the former distinction between chronic anxiety and anxiety attacks. Due to its negation of all psychopathological cues and context, and while promoting an exclusively biological model of mental pathology, which is only concerned with the conscious affects and behaviour that responds to chemical molecules, the DSM has boosted the influence of medication on “psychiatric thought, especially regarding classification” and has been instrumental in the rise of pharmaco-induced psychiatry. Its responsibility for overdiagnosis and for the inflation in the number of mental illnesses is therefore obvious. Drugs are first and foremost commercial products; their producers have understood that a giant new market has been created and have forged ties with the decision makers in psychiatry. This has quickly produced many more conflicts of interest. Moreover, there is the phenomenon of “disease mongering,” i.e., the production of new illnesses. In order to market a newly developed product, or perhaps an old one which has not been sufficiently exploited, U.S. pharmaceutical laboratories launch marketing campaigns that inflate any slightly accentuated problem of human existence and attempt to either turn it into a new disease or “attach it” to an existing one, possibly by lowering the threshold of inclusion. For this purpose they recruit experts who, having carried out a number of tests, conclude that a new disease has indeed been found, suggest a partial scientific explanation and the laboratory can start marketing the therapeutic response. This is how a number of diseases were invented, over-diagnosed and above all “sponsored” –the bipolar disorder, social phobia and the attention deficit hyperactivity disorder (ADHD) (7). While scientific medicine discovers diseases, DSM-inspired psychiatry invents them. Diagnostic inflation and overdiagnosis concern the European population as well: in 2008, the European Pact for Mental Health and Well-being stated that more than 11% of the European population is affected by mental disease. This figure will no doubt increase with the arrival of the DSM-5, which has, for example, lowered the threshold for the diagnosis of the generalized anxiety disorder (GAD) from fulfilling three items to only two and moved the onset of the attention deficit hyperactivity disorder from before 7 years of age to 12 years of age. It has also provided a definition of the gambling disorder that may open the way for other future “addictions” such as to the internet, smartphones or – why not? – gardening. This is because addiction and compulsion are now barely distinguished from each other, thus potentially bringing in a large group of new patients: “An addict is someone who cannot restrain himself from doing something, or does it in a way that is considered excessively repetitive.” If we fail to react, we are thus running the risk of finding ourselves in the same situation as the United States, with nearly 10% of its adult population dependent on psychotropic medication (8) and nearly 15% of school-age children taking psychostimulants. This phenomenon of lowering the threshold of inclusion for a given mental disorder, determined by expert consensus, is not only due to the alleged influence of pharmaceutical firms; it is also a matter of an intellectual conflict of interest because these experts tend to try and broaden the scope of application of their own expertise. At the same time, they are also worried about “forgetting” certain patients, the so-called “false negatives.” This results in a trend towards expanding the populations concerned, as well as in the appearance of false epidemics: in the case of DSM-4 this happened for example with the bipolar disorder in children, with autism or hyperactivity. Finally, among the factors encouraging overdiagnosis we should also include excessive prevention, which psychiatrists sometimes confuse with prediction. It seems that in some medical domains, there has been a step back from systematic prevention, which is seen as potentially dangerous because of its invasive nature and lack of practical purpose – I am thinking specifically of the prevention of breast cancer or prostate cancer. Like its predecessors, the DSM-5 has a static and decontextualized view of mental disease, with perilous consequences. Its perspective distorts clinical studies, for example the link established between hyperactivity and future addiction or criminality – which has inspired screening campaigns focusing on very young children. Yet it is common sense that screening for TB at a kindergarten and screening for potential future criminals are ethically two very different things. Necessary psychiatric prevention must therefore be carefully regulated, so as to avoid stigmatization and abuse, and to ensure respect for individual freedom. In this regard, the DSM-5 as a tool can indeed cause significant harm.
Let us now turn to overprescription. In psychiatry, overprescription and its side effect, the overconsumption of psychotropic drugs, stems from a number of different causes. Its existence has been attested to by a collective expert report published by the French national biomedical agency INSERM in 2012: in 2010 (9), 18% of respondents aged 18-72 said they had taken at least one type of psychotropic medication in the course of the previous year. Among these many causes, I would like to focus specifically on two, which highlight the responsibility of the DSM method and the significant risk that its latest edition will worsen the situation further.
(I) The implicitly and exclusively biological model of the DSM has greatly promoted the idea that all mental diseases stem from metabolic disorders and should be remedied by medication; this particularly concerns depressions, which, as we know, are the “disease of our time.” However, in reality, mild and moderate depressions, which constitute the great majority of depressive states leading to a consultation with a GP or a specialist, respond very favourably to psychotherapy because they are in fact what used to be called reactive and neurotic depressions. The latter are not recognized by the DSM-5, which groups them all under the rubric of “major depressive episode.” Still, future practitioners and post-graduate students should again be taught psychopathology, which distinguishes between the different psychical structures – such as neurosis and psychosis – while the DSM-derived “guidelines” for practitioners, together with the marketing pressure exerted on them by the sales representatives of pharmaceutical companies, influence their opinion unilaterally, which results in quasi-automatic prescription. An experiment carried out in the Netherlands asked general practitioners, who were working with patients that could otherwise fit the DSM criteria for depression, to restrain from diagnosis and prescription for a period of time. The reduction in prescription was spectacular and had no harmful effect on the patients’ health (see the above-cited article by Godon).
(II) The DSM and especially the DSM-5 decontextualize the observed disorders to a significant degree, especially when dealing with children. This means that the psychic suffering of a child or adolescent is reduced to the diagnosis of one or several disorders, which are considered – with no scientific proof – to be dysfunctions of the brain. Because context is not taken seriously into consideration, the child is often turned into a psychiatric patient and stigmatized, despite the fact that the condition frequently stems from the subject’s pathogenic interaction with either the family or the educational or pedagogical environment, or even with all of them together, rather than simply from a brain dysfunction. Hence the temptation and error, potentially harmful to the child’s future, of “treating” complex and multifactorial problems (psychological, biological, pedagogical, etc.) by simply prescribing a drug. The use of psychotropic medication in children should always be carefully justified; as in the case of an adult patient, it should always be communicated to the child and accompanied by psychological support for both the patient and those around him. A potential risk also lies in the physiological status of the child as a growing organism with a developing brain. In addition, children have a high level of symptomatic mobility and often exhibit a whole number of different pathological conditions described by the DSM-5. This frequently results in the prescription of multiple drugs – one for each disorder – and an accumulation of side effects. It appears that the American Psychiatric Association has become aware of this problem and has reacted by a campaign that recommends physicians to act carefully and “choose wisely” when prescribing psychotropic medication to children (10). However, this is only a sideways move in relation to the DSM-5. Independently of its effects on prescription, the de-contextualized view of the DSM-5 represents a danger in itself – the growing disinterest of clinicians in providing psychological and social care for their patients. The medical response is seen as primarily biological and thus in some sense technological, while in reality a great number of recoveries from chronic conditions have to do with modifications in the patient’s life, with their becoming involved in social activities and projects, and with the support provided by a therapist, rather than with medication itself – indeed sometimes they occur after medication has been stopped. This is evidenced by the countless cases of former patients who have been “cured” and join the groups of “psychiatric survivors” or similar organizations, or become involved in anti-psychiatry.
Let me now turn to my final point: the DSM-5 is dangerous because it leads to over-medicalization.
How is this possible? A few telling examples may help shed light on the issue: if a person who is grieving the loss of a loved one presents, after two weeks, psychomotor retardation, negative thinking, insomnia and sadness, he will no longer be considered as undergoing normal bereavement but as suffering from a major depressive state, i.e. a pathology which might require the prescription of psychotropic drugs. This can of course prove to be very harmful because it prevents the elaboration of the loss, impedes the spontaneous work of mourning, with both mid-term and long-term effects on the psyche. Another example: an elderly person with slight problems of memory could easily be diagnosed with minor cognitive disorder. In the name of the prevention of Alzheimer’s disease, such diagnosis could trigger a series of costly and useless tests, as well as the prescription of ineffective medication. I could also mention gluttony, which now becomes hyperphagia; the fact that a difficult child may quickly be diagnosed with a pathology; or that a subject with a somatic disease who is “excessively concerned” with his illness becomes a psychiatric patient suffering from the “somatic symptom disorder.”
I have already mentioned the influence psychotropic medication has had on psychiatry – a phenomenon sometimes called “pharmaceutically induced psychiatry”. Let me clarify this. If psychiatric diagnosis remains completely a matter of clinical practice and the aetiology of mental disease has not been explained, how should we understand the astonishing boom of the pharmaceutical industry, which has been bringing out new psychotropic drugs with great regularity, despite the fact that no biological markers for these diseases have been found? Before we can answer this question, I must say a few things about the originality of pharmacological research in the field of psychotropic drugs. In the case of any disease, testing the effectiveness of a new drug usually requires multiple stages of clinical trials, where the first stage only tests on cells, tissues or living animals. In general, there is a ‘reliable witness’ corresponding to the illness, for example the virus or bacteria that is its causal agent. If we want to test a drug, we inject the animal with the causal agent, the so-called reliable witness, to evaluate the drug’s effectiveness. In the field of mental illness, there is no such reliable witness because the aetiology of the disease is not known. Nevertheless, new drugs are tested on animals; the point is of course not to make a lab mouse depressed, but to try to reproduce and compare the drug-induced effects with the effects of antidepressants on humans. From the perspective of pharmaceutical research this empirical method of “pharmacological induction” is nonetheless very fruitful and as a consequence, the market has been flooded with new chemical substances. What we see here is a skilfully maintained confusion between the research dynamics and true scientific breakthroughs, which are in fact inexistent or only minimal. This approach, specific to psychopharmacological research, allows to precisely identify the behaviours and emotions affected by these chemicals and to test for small differences between the latter’s effects. It is absolutely true that this method provides us with useful information about neurotransmitters. However, its aim has been to refine the tools for distinguishing between psychotropic drugs, in order to find other, more effective ones, and ultimately to draw a kind of atlas, an exploration of all human behaviour, of all human emotions, which will be targeted by future medications. As a result, psychiatry becomes concerned exclusively with observable behaviour, with only the obvious and conscious emotions. And this is precisely what the DSM-5 has to offer, with two consequences – (A) a synergy between pharmacological research and the development of the Manual’s nosographic categories; and (B) a growing pathologization of human emotions and behaviour, as a reaction to the ever-expanding field of action of psychotropic drugs. There is no need to invoke potential conflicts of interests or any kind of “Big Pharma” conspiracy; this synergy can be explained simply as a domino effect. Starting with the DSM-3 and its focus on behavioural symptoms, psychiatry has entered a new historical era of pharmaco-induced clinical psychiatry, which follows the rhythms of the marketization of products that psychotropic drugs represent to the booming pharmaceutical industry. New medication leads to the creation of new diagnoses or the modification of existing ones. Consequently, it matters very little whether the DSM is clinically binding for European psychiatrists or not because in our liberal and globalized world both the spirit of the DSM and the pharmaco-induced model of clinical psychiatry are quickly becoming universal.
Before I finish, I would like to mention one type of danger that is specific to the DSM-5. I am referring to the medico-legal risks: if the DSM-5 is used as a legally binding reference in courts, it can lead to a situation where certain types of criminal behaviour will be considered effects of a mental disease, thus removing criminal responsibility. I am specifically thinking of sexual crimes because the DSM-5 does not make a clear distinction between, for example, a sex offender and a patient suffering from a compulsion. This is indeed an important social issue; however, the DSM has so far not “invaded” European courts, with only a few exceptions such as the Swiss Confederation.
In conclusion, I would like to stress that mental disease is not a myth. However, the way mental pathologies are classified and named, such as in the DSM-5, depends much more on social constructions influenced by the changing fashions, by the values of our time, by the power relationships between the intervening parties — medical professionals, experts, representatives of the industry, the state, third-party payers, service users, and so on and so forth — then on the scientific method.
It was this absence of scientific reasoning that led Thomas Insel, the director of the U.S. National Institute of Mental Health (NIMH), the world’s largest institute for psychiatric research, to abandon the DSM-5 and search for the biological markers of mental disease in the brain. I have shown that the DSM-5 does indeed pose risks to public heath; it serves as a reference for the clinical trials based on which the European Medicines Agency authorizes the marketing of psychotropic drugs, as well as a reference tool for research and epidemiology, despite the fact that it is not legally binding, contrary to the WHO’s own classification, the CIM 10. In the future, these dangers will surely become more concrete, especially with the appearance of new false epidemics and the rise in the costs of care necessary to treat the wrongly diagnosed patients — at the expense of those who would legitimately benefit from resources that are, after all, far from unlimited.
 François Gonon : Quel avenir pour les classifications des maladies mentales? Une synthèse des critiques anglo-saxonnes les plus récentes, L’Information psychiatrique Volume 89, Numéro 4, Avril 2013
 Royal college of psychiatrists, changing minds voir RCPsych website.
 Roy Porter :Enligthment : Britain and the creation of the modern world, Ed : Penguin 2001
 Michel Foucault : Histoire de la folie, à l’âge classique : Ed : Gallimard 1972
 United nations General Assembly 01 février 2013 Human Rights Council 22nd session Report of the special Rapporteur on torture
 Nancy Andreasen : How schizophrenia affects our brain in now.uiowa.edu/2013/08/ how-schizophrenia-affects-brain
 Michel Minard : LE DSM-ROI Ed : Erès collection”Des travaux et des jours” 2013 P. 298
 Allen Frances : Saving Normal, Ed : William Morrow 29/04/2013 p.50 à 89
 Expertise collective de l’Inserm de 2012 : Médicaments psychotropes : consommations et pharmacodépendances