The Diagnostic and Statistical Manual of Mental Disorders (DSM) is one of the two standard classification systems of mental disorders used by mental health professionals. DSM originated in 1952 (DSM-I); the other widely used system—the International Statistical Classification of Diseases and Related Health Problems (ICD)—for the first time included a section on mental disorders in 1949 (ICD-6). Both the American Psychiatric Association (APA) and the World Health Organization (WHO) are currently working on revisions of the respective classification systems. DSM-VFootnote 1 (http://www.dsm5.org) and ICD-11 (http://www.who.int/classifications/icd/ICDRevision) are scheduled for publication in May 2013 and in 2015, respectively. They will replace DSM-IV and ICD-10 which were introduced in 1994 (the “Text Revision” of DSM-IV was published in 2000) and 1992. In the United States, DSM is used for both clinical and research purposes; outside the USA, clinically orientated research is frequently based on DSM, one of the major reasons being that many research journals require studies to be based on the DSM classification. DSM not only influences how mental health specialists diagnose and treat their patients but also sways how US insurance companies decide which disorders to cover, how pharmaceutical companies design clinical trials and how funding agencies decide which research to fund [1].

The cross-talk between DSM and ICD implies that DSM-V will be of substantial importance for the revisions introduced in ICD-11. Both classification systems have previously been criticized for not taking etiological factors into account. The growing insight into the multitude, complexity and heterogeneity of causes underlying psychiatric disorders, however, again leaves us with little choice other than to precisely define their core criteria based on empiric evidence. Particularly, the sparse results achieved with genome-wide association studies have undermined the optimism prevalent just a decade ago, according to which genetic and other biological findings would become increasingly important with respect to the classification of psychiatric disorders. Critical opinions such as voiced in several media interviews by Allan Frances, the chairman of the committee that was responsible for the DSM-IV, indicate that a major revision of DSM-IV is not needed at all, is a waste of money, energy and existing datasets; this simply because there is not enough new and sufficient evidence to reorganize the current classification system. The whole prospect of incorporating various kinds of cognitive and biomarkers (in particular genetic and brain imaging data) in the new classification system has failed since, despite massive very interesting and sometimes “groundbreaking news” from neuroscience and genetics, knowledge about their clinical relevance and correlates is lacunar and fragmented, and thus in essence wanting.

Due to our adherence to definition of diagnoses via psychopathological criteria we continuously need to be well aware of the limitations imposed on us by the classification systems. Our diagnostic constructs might have little to do with the underlying etiologies. Results obtained via genome-wide association and copy number variant studies indeed indicate that specific variants can predispose to more than one disorder. In this context, it is noteworthy to point out that DSM-V aims to stress dimensional aspects of psychopathology and impairment, in addition to sticking to the categorical distinctions.

A caveat to be made and a reason for great concern and worry is linked to the whole process of publishing a revision of the DSM system and its commercial and marketing implications. The APA took a rigorous approach to ensure that all members of DSM-V committees were free of conflicts of commercial interests in terms of relationships with pharmaceutical companies. However, the preparation of the new DSM classification and the publishing of the new manual in its full format and in shortened editions, and secondary literature will generate the APA hundred thousands of US dollars. Decisions on in which countries the ICD or the DSM system will be the official psychiatric nomenclature will have far-reaching implications, also in terms of financial revenues. Especially, worrisome are ideas of launching new symptom severity scales that could replace existing broadband scales such as the Achenbach scales (CBCL, TRF, YSF) and SDQ, and that will be licensed and not available in the free domain.

DSM-V will have a substantially stronger focus on development in comparison to DSM-IV. Thus, procedures for a better integration of developmental aspects including clinical presentation, natural history, developmental psychopathology and age-at-onset both within the text sections and the criteria themselves have been proposed. In addition, developmental subtypes of disorders are to be considered and evaluated [2]. Research has linked many forms of adult psychopathology to early manifestations of mental illness observable during childhood and adolescence. Today, risks for specific adult psychiatric disorders and impairment can be estimated based on the diagnosis and course of specific childhood and adolescent disorders. As a consequence, the separate DSM-IV category, Disorders Usually First Diagnosed in Infancy, Childhood or Adolescence, could be eliminated, if developmental aspects form a strong underlying theme in DSM-V via appropriate coverage in both the main texts and criteria of disorders.

However, the proposed DSM-5 category Disorders Usually First Diagnosed in Infancy, Childhood Or Adolescence continues to flourish (see Table on http://www.dsm5.org/ProposedRevisions/Pages/InfancyChildhoodAdolescence.aspx). Furthermore, there is no clear-cut underlying decision discernible as how to deal with developmental aspects. Thus, Pica, Rumination and Feeding Disorder of Infancy and Childhood are to be removed to the category eating disorders, and separation anxiety disorder to anxiety disorders. In contrast, Posttraumatic Stress Disorder in Preschool Children and Temper Dysregulation Disorder with Dysphoria (TDDD) become novel disorders within Disorders Usually First Diagnosed in Infancy, Childhood Or Adolescence.

Interestingly, the only DSM-IV disorder that in most cases is based on mutations within a single gene [3] has been proposed for removal from DSM-V. The respective work group argues that Rett’s Disorder patients often have autistic symptoms for only a brief period during early childhood, so inclusion in the autism spectrum is not appropriate for most individuals. In addition, the work group argues that the inclusion of a specific etiological entity is inappropriate. To ensure that etiology—if known—is indicated, clinicians are to be “encouraged” to utilize the specifier “Associated with Known Medical Disorder or Genetic Condition”. For biologically orientated child and adolescent psychiatrists, it is somewhat disheartening to see that the molecular elucidation of Rett’s disorder is to contribute to its removal from the classification system! It was the precise delineation of the symptoms (and DSM-IV criteria) which led to the dissection at the molecular level. Adding to the unease, is the fact that we may have to wait for a long time until “encouraged” clinicians make use of etiological specifiers.

European Child & Adolescent Psychiatry is devoting a special issue related to the preliminary DSM-5 draft revisions to the current diagnostic criteria for psychiatric diagnoses published online 2010 (http://www.dsm5.org/Pages/Default.aspx). Novel updates are being introduced into the DSM-5 website, illustrating discussions as how to optimally capture the criteria are ongoing. The articles in this issue deal with several pertinent aspects pertaining to the proposed criteria of specific disorders particularly relevant to child and adolescent psychiatry. We have asked all contributors to pragmatically focus on clinical and developmental considerations. In order to provide the reader with a concise synopsis of disorder specific issues we asked the authors to limit both the length of the articles and the number of references, respectively. The editors and the authors are well aware of the fact that the contents of the articles do not represent any official statements, instead the contributions are based on the experts’ perceptions of the major issues pertaining to the proposed preliminary criteria. As such, it is up to the reader to carefully judge to what extent the comments capture the core issues. It is our aim to foster discussion of the criteria, which we will all depend on as of 2013.

Stringaris [4] discusses what is currently still named TDDD, but what will be caught under a new name in the final version of DSM-V. Irritability appears to have the properties of a dimension in psychopathology: it cuts across a range of psychiatric disorders including both internalizing (depression, generalized anxiety disorder and dysthymia) and disruptive disorders (oppositional defiant disorder). Potentially due to these overarching aspects, the respective working group may not have placed this disorder into the category Mood Disorders. Stringaris reviews the relationship of irritability with psychopathology and adverse outcomes and addresses the position of irritability in psychiatric nosology. The proposed introduction of TDDD must also be viewed in the context of the ongoing controversy pertaining to pediatric bipolar disorder; Stringaris suggests that TDDD may be seen as an attempt to defend against the over-diagnosis of bipolar disorder in the USA. He argues that research into measurement, phenotypic refinement, developmental aspects and genetics is required. Stringaris cautions that the TDDD label is not entirely without problems in that the word “temper” can mislead people into believing that temperament is referred to and that psychiatrists are giving diagnostic labels to temperamental variation.

Early intervention has been advocated in and outside of psychiatry to prevent the development of full-blown disorders. As such, the classificatory inclusion of diagnoses, which predict future disorders, would appear promising. In contrast to this optimistic scenario, Arango [5] addresses the risks inherent to the inclusion of the proposed novel disorder Attenuated Psychotic Symptoms Syndrome (APSS) within the category Schizophrenia and Other Psychotic Disorders. Some of the patients who in previous studies were viewed as having attenuated psychotic symptoms may actually have had a very early episode of schizophrenia. Attenuated psychotic symptoms must also be considered in developmental terms; Arango argues that the threshold for psychosis may be lower in children and, therefore, what is attenuated in adults may be fully present, although transient, in young subjects. The study of high-risk individuals is typically a domain of adult psychiatrists; child and adolescent psychiatrists are not similarly familiar with prodromal psychotic symptoms. It is crucial to realize that the specificity of psychotic symptoms to schizophrenia is considerably lower in childhood and adolescence in comparison to adulthood. Another concern is that the inclusion of APSS may lower the threshold for prescription of antipsychotics. Upon weighing the pros and cons of including this disorder in DSM-V, Arango suggests to follow the Hippocratic dictum of primum non nocere (first do no harm).

Roessner et al. [6] make a case for maintaining Tic Disorders within the category Disorders Usually First Diagnosed in Infancy, Childhood or Adolescence. The Anxiety Disorders working group also addressed Tic Disorders, thus entailing discussions as to moving this group of disorders into Anxiety Disorders. The reason for urging maintenance in the original category is that Tic Disorders bear no direct relationship with Anxiety Disorders; if the category Disorders Usually First Diagnosed in Infancy, Childhood or Adolescence is abolished, Tic Disorders should become a separate category. Roessner et al. are in favor of several of the changes in the criteria such as the unified definition of tics including the removal of the term stereotyped and the better capture of the temporal pattern of tics.

The core concept and as a result the diagnostic criteria for Attention Deficit/Hyperactivity Disorder (ADHD) have seemingly been subject to more change than for any other early onset disorder. In light of the high prevalence of this disorder and the vast amount of ADHD related research the uncertainty as to how to define this disorder is surprising; the APA has provided the scientific and clinical community with an insight into the current options for classification of ADHD. These pertain to the (a) general structure, (b) attention deficit without hyperactivity, (c) change in number, content or distribution of criteria, (d) age of onset of symptoms/impairment and change from impairment to symptoms, (e) adult ADHD, (f) ascertainment of cross-situationality, (g) inclusion and exclusion criteria, and (h) elaboration of criteria descriptions (http://www.dsm5.org/Proposed%20Revision%20Attachments/APA%20Options%20for%20ADHD.pdf). The working group has updated the original draft version in May. Coghill and Seth [7] propose to refrain from introducing changes that have not been subjected to extensive testing in field trials. They are concerned that such changes would signal that even the experts cannot agree as to what constitutes ADHD, thus increasing confusion both among clinicians and the lay public.

Von Gontard [8] argues that the proposed DSM-5 criteria for elimination disorders do not consider a wealth of novel research in general paediatrics, paediatric nephrology, urology and gastroenterology. New international classification systems reflecting this progress have been proposed by the International Children’s Continence Society (ICCS) for urinary and by the ROME-III Classification Group for fecal incontinence. For enuresis, von Gontard perceives the lack of defined subgroups as the major drawback. The proposed DSM-5 criteria only distinguish between daytime, nighttime and combined subtypes; based on the comparison with ICCS, the proposed subtypes refer to limited diagnostic possibilities only and are thus neither up-to-date nor do they reflect current research insights. While the proposed criteria for encopresis, which were left unchanged from DSM-IV-TR, are deemed somewhat better than those for enuresis, von Gontard concludes by stating that the proposed criteria for elimination disorders “lag behind basic and clinical research findings by decades”.

The diagnostic criteria for conduct disorder are to remain unchanged. However, an additional specifier for Callous and Unemotional Traits in Conduct Disorder has been proposed; its historical perspective, research basis, and development is extensively discussed by Frick and Moffitt [9] in a DSM-5 online article. Scheepers et al. [10] summarize the background information provided by the two aforementioned authors and comments on the proposed criteria including potential harmful labeling effects, neurobiological influences, and concern about the appropriateness of making the specifier contingent on Conduct Disorder, rather than maintaining some independence between callous-unemotional traits and antisocial behavior. Scheepers welcomes the specifier because it can help to specify diagnosis and further investigate etiology and treatment possibilities in Conduct Disorder.

Knoll et al. [11] expand on previous work of Hebebrand and coworkers pertaining to the criteria for Anorexia Nervosa. Based on their criticism, the DSM-IV terms “refusal” (to maintain body weight at or above a minimally normal weight) and “denial” (of the seriousness of the current low body weight) are no longer used to define this eating disorder. However, other important issues merit consideration: Due to the perception that weight phobia and fear of gaining weight form the core psychopathological features of Anorexia Nervosa, many children and adolescents will not qualify for this diagnosis, because they at least initially do not present with these symptoms. The proposed criteria would thus perpetuate the frequent need to resort to a diagnosis of an Eating Disorder Not Otherwise Specified (EDNOS) despite a phenotype that in all other aspects resembles Anorexia nervosa. The unnecessary repetition of “low (body) weight” in the three proposed criteria precludes the diagnosis of EDNOS in subjects who have a body weight above the cutoff for “significantly low weight”. Hebebrand and coworkers argue that the core symptomatology of Anorexia Nervosa rests on the intertwining of the primary behaviors with the psychological and physical consequences of starvation. They propose an alternative set of criteria that in their opinion better address this core phenotype. These criteria include both precise cutoffs for underweight based on the body mass index (BMI) and BMI age centiles and the requirement of symptoms of starvation.

Finally, Wilkinson and Goodyer [12] comment on the proposed new category of non-suicidal self-injury (NSSI). Impulsive and/or repeated self-injury is relatively common among adolescents, which currently can only be classified as a symptom of borderline personality disorder. This category of personality disorder, however, will not always be appropriate, especially in younger adolescents whose personality is still developing. NSSI mostly has no suicidal intent but is associated with reducing distressing and painful affect. Adolescents with NSSI, however, are at increased risks for later suicidal behaviours, and should be carefully assessed for suicidality. Wilkinson and Goodyear welcome the new category of NSSI, but would have liked to see the impairment criterion better phrased if it stated that self-injury is associated with, rather than causal for, intense distress.

Johannes Hebebrand

Jan Buitelaar