The failure of the schizophrenia concept and the argument for its replacement by hebephrenia: applying the medical model for disease recognition

Abstract

The present DSM and ICD delineations of schizophrenia do not identify homogeneous populations, and patients with different presentations satisfy the official criteria. Treatment response, illness course, and biological findings vary widely, indicating heterogeneity and not a common pathophysiology. The DSM/ICD construct of schizophrenia does not meet the standard of the medical model. This model for disease recognition delineates syndromes and then attempts to validate them by course and prognosis, response to treatment, and laboratory tests that ultimately lead to a clear picture of the pathophysiology and its etiology. This model has been successfully employed for centuries (1). We examine the historical record and empirical data for schizophrenia from this perspective and find that hebephrenia is a more homogeneous construct with distinctive and reliably identified clinical features and better fits the application of the medical model. Much of the literature on schizophrenia is based on samples that are heterogeneous in constitution. The non-specific findings are best summarized by Tandon et al.: Of 77 variables considered ‘important’ in schizophrenia, only 23 met the standard of support from ‘independent studies with consistent replication and no contradictory findings’ (2). The annual incidence and prevalence of schizophrenia varies fivefold across countries. The condition is highly heritable but with great genetic variability. Several environmental factors of ‘small effect’ play a role. Persons with the diagnosis have reduced brain volume with larger lateral and third ventricles than non-ill comparison groups. Dopamine agonists exacerbate and dopamine D2 antagonists ameliorate hallucinations and delusions. The heterogeneity in neurobiology, clinical manifestations, severity, course, and treatment response is great. No single feature is pathognomonic, and nosologic ‘boundaries between schizophrenia and other psychiatric disorders are indistinct’. The diagnosis is ‘characterized by an admixture of positive, negative, cognitive, and mood symptoms’. Generalized intellectual impairments occur. ‘There is a higher occurrence of obesity and cardiovascular disease’, an ‘increased prevalence of cigarette smoking and other substance use disorders’, and ‘increased suicidality’. The onset of psychotic symptoms usually occurs during adolescence or early childhood and is earlier in men. Mortality risk is doubled. This olio of non-specific observations does not support a syndrome or a disease. Genetic studies yield few positive results despite four decades of effort (3). The most accepted haplotype or allele for the cell protein dysbindin accounts for only 2% of the variance in individual differences in illness presentation (4). In these investigations, the laboratory procedures are better detailed and refined than are the descriptions of the patients studied. Unless a better defined phenotype is delineated, progress is unlikely in identifying an endophenotype or gene for schizophrenia. Traditional subtyping of schizophrenia into paranoid, disorganized, catatonic, undifferentiated, and residual forms focusing on the most prominent feature at the time of examination has not been productive (5, 6). Such sorting is mostly ignored in clinical practice and research studies. Many patients with manic depression are paranoid, catatonic, and behaviourally disorganized, but few exhibit the signs of emotional blunting, avolition, and formal thought disorder, the features that better define a syndrome and prognosis (7–12). The positive–negative symptom dichotomy also yields muddled results (11–13). Positive and negative features are frequently mixed. Patients identified by the presence of positive features also have negative features, resulting in substantial overlap. Formal thought disorder (considered a positive feature) is commonly associated with negative features, guaranteeing a mixture of symptoms in many patients (11). The suggestion of dividing schizophrenia into paranoid and non-paranoid forms, the latter to include catatonia and hebephrenia, has not been adopted (14, 15). Present DSM diagnostic criteria guarantee sample heterogeneity. Should most patients with hallucinations and delusions and without mood disorder or identifiable neurologic disease be labeled schizophrenic? Is schizophrenia a developmental disorder emerging early in life before psychosis or a disease afflicting any age group? Is a residual decline in function always present or can the illness remit? Is a 17-year-old isolating person with long-standing problems of emotional expression, volition, and social and motor awkwardness suffering from the same illness as a 50-year-old person with adequate premorbid functioning, if both meet the cross-sectional criteria? A person with preserved personality and normal emotional expression but experiencing persistent auditory hallucinations of first rank (e.g. voices commenting) and a person with catatonia and delusions of grandeur or disorganized speech both meet the criteria. Do they suffer from the same illness? Any two features permit the DSM diagnosis: hallucinations, delusions, ‘disorganized’ speech, ‘disorganized’ behaviour or catatonia, and negative symptoms. A ‘bizarre’ delusion (e.g. being controlled by electromagnetic waves) or a hallucination of sustained voices discussing or commenting in the third person about the patient is each sufficient alone; but reliance on a single feature to diagnose schizophrenia harkens back to Kurt Schneider’s notion of first rank symptoms (16). These features, while characteristic, are not pathognomonic of schizophrenia, even in the absence of coarse brain disease, nor do they predict treatment response or long-term prognosis (11, 17, 18). These weaknesses in the present formulation of schizophrenia led to our review. Applying general paresis as a model for a single disease frequently expressed as several syndromes, Wilhelm Griesinger (1817–1868) championed the mid-nineteenth century notion of what was later called a ‘unitary psychosis’– different syndromes considered stages of a single disease rather than expressions of different pathophysiologies (19). Sufferers were believed to pass through stages of melancholia, mania, and amentia (delirium), ending in dementia, although it was understood that some patients experienced several forms, while others experienced only one (20). With only one form of psychosis to consider, the profession’s energies were directed at determining etiology, not classification. Karl Ludwig Kahlbaum (1828–1899) rejected the unitary model and re-focused attention on delineating specific syndromes, seeking to link psychiatric disorders to biological turning points in life (21–23). ‘Paraphrenia hebetica’ was identified as the psychosis of youth and adolescence. He coined the term ‘hebephrenia’. Kahlbaum classified behaviour disorders by symptom patterns and illness course in 1863. Basing his studies on the 19th century’s tripartite image of the mind consisting of will, emotion, and intellect, he recognized an idiopathic progressively deteriorating condition affecting all spheres that became Kraepelin’s model for dementia praecox. He defined a postpubescent circumscribed illness affecting only emotion that is recognized in today’s mood disorders. Melancholia was included in this category, but he used the term ‘dysthymia’ for it. The concept of ‘cyclothymia’ was introduced as a low-grade form of what became ‘manic-depressive illness’. His disorders of intellect (e.g. paranoia, dementia paranoides) are today’s delusional disorder, and his disorders of will are exemplified by catatonia (23). At the Reimer Sanitarium in Görlitz, Kahlbaum was joined in 1866 by Ewald Hecker, a junior clinician (21, 22). In 1874, Kahlbaum published his monograph on catatonia, separating catatonia and hebephrenia (24). He presented catatonia’s distinctive clinical features, course, and a variety of non-specific brain autopsy findings as validation for catatonia being a discrete disease. Catatonia is now recognized as a syndrome associated with many conditions (25). A chronology is cited in Table 1. In 1871, Ewald Hecker (1843–1909) described ‘hebephrenia’ as a discrete illness by its symptoms and course, acknowledging Kahlbaum as the originator of the concept (26, 27). Hecker wrote: ‘Of course, not all mental illness appearing at puberty show the same development. Nearly all the clinical forms (mania, melancholia, etc.) can be found in this age group without appearing much different than they do at other ages. Hebephrenia, however, stands out as possessing both a specific course and a distinct symptomatology’. After presenting a typical case history, he concluded that ‘Hebephrenia… is a disease that always rises in connection with the development of puberty’ (27). Among his patients, the onset was between ages 18 and 22, the course distinctive, always ending in ‘hebephrenic dementia’; the disease process ‘limits further intellectual and emotional development and produces a particular form of mental disability’. The disease began with symptoms of melancholia but then psychosis set in. Amidst the symptoms of melancholia, ‘these patients often exhibit an irrepressible impulse to laugh and tell silly jokes’. [Witzelsucht] After discharge from the hospital, many subjects became vagabonds. Some philosophized grandly about life and science. Non-sequiturs and poorly constructed logical sentences crept into their speech. Some spoke and wrote in strange jargons. While all the symptoms would not be present in every patient, ‘the characteristic course that is invariably present, the early onset that is unmistakable, and the peculiarly silly form of dementia together make for a secure delineation of this type of illness’. As the illness advanced, the silliness was replaced by ‘hebephrenic dementia’, a state that did not approach the dementia of neurosyphilis (which was well recognized), but which corresponded to ‘an intermediate level of mental deterioration’. Hallucinations emerged. Later, listless apathy replaced the silliness. The distinctive and historically new element in Hecker’s description was the downhill course of an illness that began as melancholia, rapidly ending in intellectual deterioration, often over a period of a few months (26, 27). This concept was quickly accepted (28–30). Emil Kraepelin (1856–1926) adopted Kahlbaum’s classification system: ‘I got the starting point of the line of thought which in 1896 led to dementia praecox being regarded as a distinct disease, on the one hand from the overpowering impression of the states of dementia quite similar to each other which developed from the most varied initial clinical symptoms, on the other hand from the experience connected with the observations of Hecker that these peculiar dementias seemed to stand in near relation to the period of youth’ (31). Also: ‘I kept Kahlbaum’s and Hecker’s ideas in mind and tried to collect those cases, which inclined toward dementia as ‘mental degeneration processes’. Apart from Kahlbaum’s catatonia, I differentiated between dementia praecox, which essentially corresponded with hebephrenia, [our emphasis] and dementia paranoides with hallucinations, which quickly developed into mental deficiency’ (32, 33). Relying heavily on illness course, Kraepelin conflated the syndromes of catatonia, paranoia, and hebephrenia into the single condition: ‘I was forced to realize that in a frighteningly large number of patients, who at first seemed to have the syndrome of mania, melancholia, insanity, amentia or madness, the syndrome changed fairly quickly into a typical progressive dementia and in spite of some differences, the syndromes became increasingly similar. I soon realized that the abnormalities at the beginning of the disease had no decisive importance compared to the course of the illness’ (32). Kraepelin’s interpretation of dementia as a necessary outcome compounded a faulty reliance on catatonia and hebephrenia as one disorder. Many patients with severe mood disorder appear demented, and a large number become chronically dysfunctional when inadequately treated. Without effective treatments, severity and chronicity were common in Kraepelin’s experience, encouraging him to incorporate manic and melancholic patients, particularly those with catatonia, into his notion of dementia praecox. Outcomes for the different variations of dementia praecox were grim: ‘I must consider excluding the possibility of any recovery from this illness’. He reserved the label of hebephrenia only for those cases in which the presence of ‘agitation’ creates ‘the more unfavorable kinds of prognoses’ (34). The image shifted in successive editions of his textbook. In the sixth edition, Kraepelin divided dementia praecox into hebephrenic, catatonic, and paranoid forms. He retained Hecker’s ideas of hebephrenia (35). In 1899, he included délire chronique (36) within dementia praecox. In the seventh edition (1904), he added all disorders that led to chronic behavioural and intellectual impairment (37). Otto Diem’s dementia simplex was included (38), but the paraphrenias were separated by delusions without avolition or loss of emotional expression. In the eighth edition (1913), Kraepelin introduced the unfortunate term ‘silly dementia’ (läppische Verblödung) to characterize hebephrenia. The term has misled the profession, as ‘silly’ did not mean jocularity but rather a lapse into immaturity in socialization and thinking with blunting of emotional expression (this was also Hecker’s understanding of the term). Kraepelin added descriptions of psychosis, highly changeable moods, isolation in relations with others, and ‘irrational and confused’ actions. Almost all ended in dementia with few recoveries. The patients with silly dementia (hebephrenia) amounted to thirteen per cent of the cases of dementia praecox (39). Among other patients with dementia praecox, Kraepelin recognized that some recovered (31). The validity of Kraepelin’s concept of dementia praecox rested on both catatonia and hebephrenia being aspects of a single disease with a deteriorating course. By the end of his career, he had conceptually merged dementias occurring before age 50 and all forms of mania, melancholia, and circular mood disorders into the two ‘functional’ psychoses of dementia praecox and manic-depressive insanity. This formulation became the bedrock of the ICD and DSM classifications. Its weak logic (all conditions with the same longitudinal course have similar etiology) and faulty construct (the effect of the illness on the tripartite mind determining grouping) shape present-day efforts to delineate psychiatric illness. Eugen Bleuler (1857–1939), professor of psychiatry in Zurich, transformed Kraepelin’s concept of dementia praecox in 1908 renaming it ‘schizophrenia’ (40). Fundamental symptoms were identifiable in all the patients, with fluctuating accessory secondary symptoms in some. Kraepelin saw one disease. Bleuler acknowledged the possibility of several disorders, with ‘hebephrenia’ as the principal form (41). In 1911, Bleuler argued that dementia was too narrow a view of the condition. He rejected the notion of ‘pubertal insanity’ and said hebephrenia could occur at any age (42). By 1916, Bleuler rejected silly adolescent behaviour as a criterion for the diagnosis and dismissed the concept of adolescent insanity inevitably ending in terminal dementia. He retained the term ‘hebephrenia’ nevertheless, ‘although it no longer applies to the current concept’ (43). Bleuler accepted and expanded the dementia praecox notion. He used the never validated idea of fundamental and accessory features to characterize the illness. His terminology and constructs were easily misunderstood. And he had the mischance to write in German just before two world wars that resulted in things German being shunned and the emergence of English as the dominant language of science and medicine. His writings were not available in English until 1950 and are not commonly studied in psychiatric training centers. The heterogeneity of samples identified by Bleuler’s approach led to efforts to separate ‘good’ from ‘bad’ prognosis schizophrenic patients (44), the construction of the hybrid schizoaffective disorder (45), and proposing schizophreniform disorder (46) and reactive psychosis (47) each of ‘psychogenic origin’ with good outcomes. These constructs are now conflated as brief psychotic disorder. None has been empirically validated. Recognizing the heterogeneity among patients with schizophrenia, some writers sought to recapture the image of hebephrenia. In the 1920s, Karl Kleist, professor of psychiatry at the University of Frankfurt, found hebephrenia to be among schizophrenia’s core symptoms. Kleist wrote of ‘affective atrophy’ in hebephrenia, an illness whose cardinal symptom is affective flattening, not silliness. Schizophrenia was not a group of syndromes but a single disease, the ‘process of progressive dementia with deficit symptoms’. Hebephrenia and catatonia were its core symptoms (48). Kleist’s student Karl Leonhard, professor of psychiatry at the Charité Hospital in East-Berlin, regarded ‘affective flattening’ and being ‘affectively deadened’ as the prime characteristic of hebephrenia, one of his ‘systematic’ psychoses (meaning non-fluctuating psychoses)’ (49, 50). While Leonhard’s overall schema has not been accepted, his description of systematic psychosis of hebephrenia is consistent with Hecker’s model. Kurt Schneider (1887–1967) proposed the notion of ‘first rank symptoms’ as specific features that, in the absence of coarse brain disease, were pathognomonic of ‘true’ schizophrenia with its poor prognosis (51). Although first rank symptoms are found in mania and melancholia (9, 16, 52, 53) as well as in schizophrenia, they form the basis for the DSM ‘criterion A’ for schizophrenia. Their occurrence is de facto considered pathognomonic as the schizophrenia diagnosis is permitted if the patient only exhibits the hearing of sustained voices discussing or commenting in the third person about the patient or has a ‘bizarre’ delusion (e.g. believing a transmitter has been placed in a tooth that reveals one’s thoughts). The features of hebephrenia are displayed in Table 2. Individual features are not independent. They have been historically associated with hebephrenia, and the motor and cognitive features are linked to brain frontal circuitry dysfunction. Loss of emotional expression, avolition, and indifference are classic features of the frontal lobe apathetic syndrome (54, 55). Executive function difficulties are an accepted aspect of frontal circuitry, the brain region particularly vulnerable to such environmental impacts as maternal starvation, viral infection, and anoxia, experiences reported in persons who later develop schizophrenia (56, 57). The diagnosis of hebephrenia requires four elements: i) loss of emotional expression and avolition; ii) formal thought disorder (defined as speech and language difficulties); iii) delusions of passivity or ‘complete’ auditory hallucinations; and iv) deficits in executive functioning or motor disturbances. Childhood cognitive, emotion, and neuromotor problems are commonly present. Catatonia is not a feature of the designation because it is independently defined and is effectively treated. Its presence assumes priority in treatment. Hypomania, mania, or melancholia in the psychopathology of the present illness or in the personal history directs attention to the mood disorders and precludes the diagnosis of hebephrenia. There is little historical and experimental support for schizophrenia as a single disease. The historical record, however, encourages the singular concept of hebephrenia and indicates its specific criteria. Hecker’s model of hebephrenia was of a brain illness with deficits in emotional expression, volition, motor regulation, executive functioning, speech and language, and perceptual integration. Onset in childhood with emerging neuromotor and socialization difficulties was followed in adolescence and young adulthood by hallucinations and other perceptual disturbances and delusions and then by a persistent but not progressive decline in cognitive, social, interpersonal, and employment functioning. The syndrome is recognizable today (57, 58). The hallmark negative features and formal thought disorder are found together in 60–80% of psychotic patients who have neither mood disorder nor defined neurologic disease (59–62). Carpenter distinguished such patients as having ‘deficit syndrome schizophrenia’ (63). Cluster and factor analytic studies support the descriptive validity of hebephrenia but find samples of schizophrenics to be heterogeneous (64–71). In a latent class analysis of a large proband and relative sample, Kendler et al. identify ‘classic schizophrenia’ as characterized by both ‘positive’ and ‘negative’ symptoms, chronicity, poor outcome with deterioration, no depressive or manic symptoms, low rates of marriage and employment, and a family illness pattern of little manic depression but more schizophrenia (72). Oddly, they describe a ‘hebephrenia’ group with a mixed pattern of psychosis, deterioration, and excited or ‘delirious’ mania. The former class better fits the hebephrenia appellation. While deficit syndrome schizophrenia offers a consistent picture, other classes of psychotic disorders exhibit class instability (73, 74). Negative features are stable over time, even between exacerbations of psychosis (75). Such patients have deficits in executive functioning (76), with prolonged poor social and work performance, and failure to respond to modern pharmacotherapy (77). Formal thought disorder, i.e. aphasic-like speech and language abnormalities, is also associated with dysfunction in frontal lobe executive functions, poor response to pharmacotherapy, chronic illness course, and similar pathology in relatives (61, 78–81). Hecker’s detailed formal thought disorder was characterized by the speech patterns of his patients. Patients with hebephrenia function poorly in their daily lives (82). Yet, many function well enough to lead independent existences, work productively, and contribute to the community. Those with negative features and early onset of illness are most likely to become chronically ill (77). Patients with poor socialization, low libido, psychosocial isolation, and reduced interests before exhibiting psychosis fail to respond to pharmacotherapy, have persistent symptoms, and chronically function poorly (77, 83). Hecker’s reports of pre-existing childhood difficulties in hebephrenia are consistent with modern reports (84–87). Bleuler described reduced emotional expression and volition, social and motor awkwardness, and isolation as prepsychotic features of schizophrenia. About half the persons diagnosed with schizophrenia exhibit these childhood behaviours (88, 89). The traits are modestly heritable (90, 91). Children who exhibit abnormal emotional expression, inappropriate social interactions, cognitive inflexibility, and neuromotor problems experience a characteristic deficit syndrome later in life. Genetic and intra-uterine factors (e.g. maternal malnutrition, influenza) (56) contribute to the condition. The more salient these factors, the more severe are the childhood abnormalities and the later psychotic episodes. While poor neuromotor and social functioning are associated with hebephrenia, the validating value of a chronic course after the first episode is unclear. When patients with schizophrenia, schizoaffective, and manic depression are compared on these criteria, no points of rarity among the groups are found (92). Long-term outcome is variable in each group with substantial chronicity in persistence of symptoms and decline in functioning in those with a deficit syndrome (92–96). Mortality rates are higher than expected (97, 98). Kraepelin’s concept of a poor prognosis dementia praecox and a good prognosis manic depression is found only at the extremes of a continuum when the present DSM criteria are applied to patient populations (99). While dementia is neither inevitable nor even the most common outcome of DSM-defined schizophrenia, cognitive difficulties are recognized early, particularly those consistent with hebephrenia. The deficits, even when mild, involve cognitive processes, with executive functions the most dramatically affected. The deficits are associated with negative features and formal thought disorder, but not with hallucinations and delusions (76, 100, 101). Some researchers propose cognitive deficits to be a diagnostic criterion for schizophrenia, but while most sufferers have cognitive problems consistent with frontal circuitry dysfunction, these are not pathognomonic (102). Synthesizing the attributes of hebephrenia to a class of patients and then predicting course has not been often assessed, but several recent efforts suggest the value of considering psychotic disorders from a non-DSM perspective. Koutsouleris et al. used MRI-derived neuroanatomic patterns (consistent with the features described in the next section) in a small sample to successfully predict transition to psychosis over a 4-year follow-up in persons with a positive family history for psychosis. Subjects had clinical features such as previous prenatal and obstetrical problems, thought interference, receptive language difficulties, and acoustic and visual perceptual disturbances (103). Ruhrmann et al. describe a clinical model for predicting future psychosis in a sample of adolescents and young adults also at high risk of psychosis. Those at greatest risk of future psychosis were likely unmarried or divorced, had obstetrical difficulties, and were said to have ‘schizotypal’ traits. These patterns are consistent with the image of hebephrenia (104). While modern treatments for schizophrenia ameliorate symptoms to some degree, none leads to remission of the illness. Both typical and atypical antipsychotic drugs and electroconvulsive therapy (ECT) relieve the ‘positive’ symptoms of hallucinations and delusions (105). No treatment effectively limits the ‘negative’ symptoms of apathy, loss of emotional expression, and avolition (106). Of the present subtypes of schizophrenia, effective treatments are known for the catatonic and paranoid types. Patients with catatonia respond remarkably well to benzodiazepines and to ECT, usually fully remitting the signs of catatonia, and often the associated signs of psychiatric illness (25). It was the reformulation of catatonia as a syndrome and not as a type of schizophrenia that encouraged clinical trials, which developed its present successful treatment algorithms (107, 108). Persecutory delusions may be ameliorated by classical antipsychotic agents and by ECT although less effectively than the relief afforded catatonia. Definitive studies of treatment response in hebephrenia are lacking, but classifying it separately would encourage such efforts. We lack a biochemical or physiological marker identifying hebephrenia. The omission likely results from such patients being lumped with others. Neuromotor difficulties consistent with abnormal development are reported for the deficit syndrome. Kraepelin described basal ganglia and cerebellar motor signs in his patients and discussed a ‘cerebellar form’ of dementia praecox (109). Cerebellar volume loss is reported in patients with schizophrenia exhibiting negative features, even in never-medicated persons (110). Cerebral cortical and subcortical abnormalities in structure and metabolic function are also reported (111–113), and basal ganglia signs are seen in never-medicated schizophrenics (114). The abnormalities are associated with the negative features of the condition (115). Schizophrenic patients with negative features also exhibit delayed and jerky pursuit eye movements, a difficulty linked to frontal circuitry disease (116). The difficulties are found early in the condition (117). The basal ganglia and cerebellum are part of a motor-cognitive system that also controls eye movements. All the motor features of the severe forms of schizophrenia can be understood as expressions of dysfunction in this system (118). Despite the successful definition of the human genome, progress is poor in identifying putative genes of vulnerability for schizophrenia. Few associational strategies yield positive results (4). The genome-wide association studies for schizophrenia reveal no clear findings (3). The results are confounded by the inclusion of patients with the schizoaffective designation (6). The difficulty in detecting the genes for schizophrenia has been attributed to the modest and polygenic heritability presumed for the condition, the putative involvement of several different chromosomal sites, and the possibility that the involved genes may be expressed as a vulnerability toward a psychotic illness but not for a specific psychotic condition (119). A more parsimonious criticism of the search for ‘the genes of schizophrenia’ is that schizophrenia, as now defined, is not a single disease. The search is analogous to the quixotic quest for the genes of mental retardation. Kraepelin reported that 70% of his patients with dementia praecox had a family history for psychosis (39), and many recent studies report increased morbid risks for psychosis among first-degree relatives of psychotic patients and adopted-away offspring of mothers diagnosed as schizophrenic. Concordance rates among mono- and dizygotic twins are elevated, consistent with a genetic transmission (6, 120). A few studies, however, identified patients by criteria specific for hebephrenia. Among those that did, the heritability for hebephrenia (identified as ‘core’ or ‘negative’ symptom schizophrenia) is weak with a possible relationship between the severest forms and mood disorder (119, 120). Only additional studies that specifically define probands by criteria for hebephrenia can resolve this issue. Unless we isolate hebephrenia as a syndrome, genetic and biological studies will remain unproductive. The historical record does not support schizophrenia as a valid diagnostic class. Psychopathologists from Kahlbaum and Kraepelin to recent writers conceived a disease that likely does not exist as a single pathophysiologic process, yet which consumes much psychiatric thinking and resources. Modern efforts to delineate the notion do not capture a homogeneous population. Clinicians and researchers are obliged to rely on vague and non-specific interview criteria to identify persons who are ill and who hallucinate or who are delusional, but who do not cluster into a clear syndrome. Biological criteria and effective treatments to verify the diagnosis are lacking. ‘Psychosis without mood disorder or identified neurologic disease’ would serve equally well as the present criteria. The only condition in the schizophrenia story that has not been rejected by evidence is hebephrenia or ‘core deficit syndrome’ schizophrenia. It is a recognizable syndrome and should be considered a replacement for the present construct of schizophrenia. Studies of the psychopathology of psychotic conditions support the hebephrenia construct, but the validity of the syndrome needs to be prospectively tested. Resurrecting hebephrenia as a subtype of schizophrenia will not suffice. Hebephrenia is not a subtype of schizophrenia. It is schizophrenia. Its characteristics are well defined and warrant its replacing the construct of schizophrenia. We are grateful to Susan Bélanger of the University of Toronto who copy edited the manuscript and organized the references.