A new study, published in Psychiatry Research, has concluded that psychiatric diagnoses are scientifically worthless as tools to identify discrete mental health disorders.
The study, led by researchers from the University of Liverpool, involved a detailed analysis of five key chapters of the latest edition of the widely used Diagnostic and Statistical Manual (DSM), on ‘schizophrenia’, ‘bipolar disorder’, ‘depressive disorders’, ‘anxiety disorders’ and ‘trauma-related disorders’.
Diagnostic manuals such as the DSM were created to provide a common diagnostic language for mental health professionals and attempt to provide a definitive list of mental health problems, including their symptoms.
The main findings of the research were:
- Psychiatric diagnoses all use different decision-making rules
- There is a huge amount of overlap in symptoms between diagnoses
- Almost all diagnoses mask the role of trauma and adverse events
- Diagnoses tell us little about the individual patient and what treatment they need
The authors conclude that diagnostic labeling represents ‘a disingenuous categorical system’.
Lead researcher Dr. Kate Allsopp, University of Liverpool, said: “Although diagnostic labels create the illusion of an explanation they are scientifically meaningless and can create stigma and prejudice.
I hope these findings will encourage mental health professionals to think beyond diagnoses and consider other explanations of mental distress, such as trauma and other adverse life experiences.”
The authors conclude that diagnostic labelling represents ‘a disingenuous categorical system’. The image is in the public domain.
Professor Peter Kinderman, University of Liverpool, said:
“This study provides yet more evidence that the biomedical diagnostic approach in psychiatry is not fit for purpose.
Diagnoses frequently and uncritically reported as ‘real illnesses’ are in fact made on the basis of internally inconsistent, confused and contradictory patterns of largely arbitrary criteria.
The diagnostic system wrongly assumes that all distress results from disorder, and relies heavily on subjective judgments about what is normal.”
Professor John Read, University of East London, said:
“Perhaps it is time we stopped pretending that medical-sounding labels contribute anything to our understanding of the complex causes of human distress or of what kind of help we need when distressed.”
THE CONCEPT OF SYNDROME IN PSYCHIATRY
The concept of syndrome is relatively recent. Most of the schools of medicine of the ancient world considered every single symptom and sign as a separate disease itself .
In 1742, Sydenham was the first scientist to clearly formulate the minimum requirements needed to define a clinical condition as a syndrome.
In his definition, a syndrome consists of several interrelated symptoms showing a stable, characteristic structure and a peculiar prognosis .
Furthermore, patients affected by a specific syndrome should share a sufficiently pathognomonic cluster of symptoms that allows clinicians to distinguish them from other patients suffering from different syndromes.
Such fundamental elements of discontinuity among various syndromes were, at first, better conceptualized by Sneath , who introduced the term of “point of rarity” to refer to precise clinical boundaries among disorders, and were later revised by Kendell and Jablensky , who preferred the concept of “zone of rarity” separating psychiatric syndromes.
The idea was that if a syndrome corresponds to a natural entity, than we should find a natural boundary or a discontinuity between this condition and its clinical “neighbours” ; in other terms, mixed conditions can exist, but they have to be less common than the pure syndromal form.
Several authors [i.e. 1, 2, 6] have investigated whether such a theoretical construct could be applied to the field of mental disorders. In spite of the improvement in this field, the validity of the actual classifications (ICD-10 , DSM-IV-TR ) still remains a matter of debate.
Although formal criteria to establish the validity of psychiatric diagnoses have been progressively proposed (i.e., descriptive features, sociodemographic factors, laboratory markers, exclusion criteria, neurobiologic and genetic validators, prognostic characteristics, level of functional and social disability, response to therapy, statistical deviation from the healthy condition, etc.) [9–11], in a clinical framework some of the conditions that are currently classified as psychiatric disorders still appear as vague and confusing (rather than discrete) entities. In particular, most of them show weak diagnostic criteria (low association between diagnostic symptoms and actual diagnosis), unspecific manifestations (poorly pathognomonic symptoms), and widespread onset (high comorbidity).
AN EXAMPLE OF THE UNSTABLE DIAGNOSTIC CRITERIA IN PSYCHIATRY: THE GENERALIZED ANXIETY DISORDER
According to the mentioned Sydenham’s definition of syndrome, a first anomaly has to be taken into account: several diagnostic categories changed their criteria in different DSM editions and some of them were abolished, merged with other conditions, or divided into two or more entities.
GAD was first introduced as a unique diagnosis in the third edition of the DSM (DSM-III) , but it was most often used as a residual diagnosis for individuals who did not meet diagnostic criteria for another anxiety disorder .
As a consequence, the DSM-IV-TR  considered GAD as an independent diagnosis among the anxiety disorders, characterized by “excessive, uncontrollable worry about a variety of topics that occurs more days than not for a period of at least six months”.
The worry causes distress and/or functional impairment, and is associated with at least three of the following features: restlessness or feeling keyed up or on edge, being easily fatigued, difficulty concentrating or mind going blank, irritability, muscle tension, and sleep disturbance (difficulty falling or staying asleep, or restless unsatisfying sleep) .
Indeed, four out of the six associated physical symptoms of GAD (i.e. restlessness, fatigue, difficulty concentrating, sleep difficulties, obsessive rumination, somatization) are also part of the diagnostic criteria for Major Depressive Disorder [16, 17].
Furthermore, symptoms required for GAD diagnosis are also present in other anxiety disorders, as closer overlaps probably exist between GAD and Panic Disorder or Social Anxiety Disorder .
In general the nosographic ambiguity – which is related to different psychiatric diagnoses – makes it very hard to define a stable constellation of interrelating symptoms and to characterize a specific population of patients with a specific diagnosis. The uncertain psychopathologic boundary of many psychiatric conditions seems to further move these disorders away from the original definition of discrete syndromes.
THE PROBLEM OF COMORBIDITY
The psychiatric comorbidity – defined as co-occurrence of two or more psychiatric diagnoses in the same patient – raised a significant concern in the debate on the categorization of psychiatric disorders.
The concept of comorbidity was first introduced by Feinstein  to denote those cases in which a ‘distinct additional clinical entity’ occurred during the clinical course of a patient having an index disease.
This concept was recently used to indicate not only those cases in which a patient receives both a psychiatric and a general medical diagnosis, but also those cases in which two or more psychiatric diagnoses are present in the same patient .
The lack of well-defined and specific symptoms for many disorders and the high rates of GAD comorbidity with other anxiety and mood disorders make it difficult for the clinician to perform a reliable differential diagnosis and to correctly identify this condition [18, 21].
Moreover, neurotic symptoms appear to be more frequent in mixed or associated forms than as parts of pure conditions. Again, the high rate for GAD comorbidity was considered to challenge the validity of this diagnosis.
SYMPTOMS OVERLAP: EXPERIMENTAL DATA FROM THE SESTO FIORENTINO STUDY
To support the concerns raised above, we performed further analyses of the data from the Sesto Fiorentino Study, a large community study, which has been described in detail elsewhere .
The main feature of this study is that psychiatric symptoms are assessed “bottom up”, i.e. regardless of the diagnosis: diagnostic categories are built up later, by computerized algorithms.
There are several advantages of focusing on symptoms rather than diagnoses. Such an approach does not depend on pre-constructed nosological theories, allows the analysis of natural aggregations of symptoms even below their diagnostic cutoff points , permits the comparison of the qualitative and quantitative aspects , and can be useful to investigate specificity of diagnostic symptoms.
We evaluated the presence of some common mood and anxiety symptoms across the most represented DSM IV diagnoses (Fig. 11).
It is clear that these symptoms occur transversally in a variety of diagnostic categories, with little variations from each other. As far as the GAD symptoms are concerned, a further analysis was performed on the data from the Sesto Fiorentino study: we tested for significant differences in each of the GAD symptom reported in Fig. (11), across diagnostic categories.
This analysis generally confirmed the lack of specificity of the reported symptoms.
These results confirmed that the symptoms which are required for a diagnosis of Major Depressive Episode or Generalized Anxiety Disorder are essentially non-specific.
Moreover, symptoms tend to be considered as “stand alone”, independently of the presence of other symptoms, whereas in the clinical practice the co-occurrence of other symptoms may consistently modify the interpretation of the clinical presentation.
Suicidal thoughts accompanied by delusions of guilt are considered potentially more dangerous than suicidal ideation plus hystrionic features.
The application of the categorical model in psychiatry is likely to be different from the one used in other medical sciences.
In general medicine, symptoms are considered either non-specific (fever or fatigue may be taken as an example), or specific (such as the aspect of blisters in HZV infection).
In medicine we do not count symptoms in order to make a diagnosis: peritonitis and flu may share some symptoms, for example fever, pallor or fatigue (non-specific), but they cannot be used to make up the diagnosis of peritonitis with a comorbid flu, and in any case if a peritonitis is present, none would assign the same value to the symptom “fatigue” and to the one “abdominal guarding exacerbated by moving peritoneum”.
DSM-III and its successors, became universally and uncritically accepted as the ultimate authority on psychopathology and diagnosis. DSM forms the basis for psychiatric teaching to both residents and undergraduates.
Although it represents an undoubted progress for reliability and communication between clinicians and researchers, the price that psychiatry had to pay for these benefits is very high: the diagnostic process lost out in terms of sophistication and specificity.
According to Andreasen , our data demonstrated that diagnostic criteria include only some characteristic symptoms of a given disorder but they were never intended to provide a comprehensive description.
Validity has been sacrificed to achieve reliability. DSM diagnoses have given researchers a common nomenclature – but probably the wrong one.
Although creating standardized diagnoses that would facilitate research was a major goal, DSM diagnoses are not useful for research because of their lack of validity.
Since the core features of mental illness were mainly unknown, both the symptom pattern and the prognostic value (e.g. duration, course and severity) of psychiatric syndromes were defined arbitrarily and conventionally.
Many examples can be mentioned to support this idea, ranging from the scarce agreement between DSM/ICD classification systems [31, 32] to the lack of validity of schizophrenia subtypes  and the problems inherent in the diagnosis of some psychiatric disorders (i.e. schizoaffective disorder) [34, 35]. T
herefore, the validity of the present nosology has never been sufficiently demonstrated. On the contrary, the premise that original DSM diagnostic categories are primarily artificial and have no confirmed value of elementary clinical entities, is usually forgotten.
This shortness of memory can push psychiatrists to think that DSM/ICD diagnoses are natural categorical entities (a phenomenon that Van Praag called “nosologomania”), thus making them accept unconditionally and uncritically the current classification system and ignore alternative models .
Many authors believe that this might represent a rather blind approach that may be either dangerous or useless in terms of application in biological and psychopharmacological research.
In fact, most of the available scientific studies are actually based on the current arbitrary classification, even if it may not match with the correspondent biological and clinical sources.
On the other hand, studies based on the so-called dimensional approach to psychopathology [37–39] show a relative aspecificity of diagnostic categories and provide robust examples of usefulness of reasoning, in terms of symptom clusters.
In our opinion, the current nosographic system prevents psychiatry from benefiting of the significant technological progress that has led the rest of medical sciences to important clinical achievements in the last 20 years.
University of Liverpool
Simon Wood – University of Liverpool
The image is in the public domain.
Original Research: Closed access
“Heterogeneity in psychiatric diagnostic classification”. Kate Allsopp, John Read, Rhiannon Corcoran, Peter Kinderman.
Psychiatry Research. doi:10.1016/j.psychres.2019.07.005