These terms summarize the different conceptualizations of the major psychoses throughout the early twentieth century. "Splitters" favour separation in terms of aetiology by way of categories. "Lumpers" postulate clustering of characteristics at the root of a predisposition to any of the major psychoses.
Wernicke (1906) proceeded conceptually from neurology, describing three functional brain systems involving the association cortex; namely, the psychomotor (awareness of own body), the psychosensory (awareness of external world), and the intrapsychic (awareness of one's personality). According to Wernicke, disturbances of these systems resulting from different aetiologies led to psychotic syndromes which could be classified as somatopsychoses, allopsychoses, and autopsychoses.
These ideas influenced Kleist (1947) and Leonhard (1979), who developed complex classifications incorporating Wernicke's ideas. Leonhard split schizophrenia into two disease groups—systematic and unsystematic, with contrasting aetiologies. The disorder was genetic in the case of the unsystematic form and developmental/environmental in the systematic form. Leonhard was the first to separate bipolar and unipolar disorders—a dichotomy which has now been adopted by the classificatory systems. There is now some evidence to support the prognostic validity of this classification (Astrup, 1979; Perris, 1974).
Kretschmer (1927) provided a prototype multidimensional classification of the major psychoses, using character trait clusters, that is, schizothyhmic, cyclothymic, and viscous, which, respectively, reflected an underlying predisposition to schizophrenia, affective psychoses, and epilepsy.
Kretschmer did not concur with the view that the psychoses were circumscribed disease phenomena, but instead held that they were episodes "rooted in the biological constitution of the individual with all possible phenomena from subclinical to florid". The genome was responsible for the "underlying all-embracing genotype"; hence, the correlation between body build and the diathesis to a particular type of psychosis—complex or mixed psychopathological pictures were the result of additive or interaction effects of inherited predispositions.
Kraepelin challenged his own dichotomous orthodoxy. He was careful to emphasize that the disease entities of dementia praecox and manic-depressive insanity were provisional. In "Patterns of mental disorder", Kraepelin (1920) described a different approach from previous views: "It is natural to turn away from arranging illnesses in orderly well-defined groups and to set ourselves instead the undoubtedly higher and more satisfying goal of understanding their structure."
Jablensky (1999) summarized Kraepelin's view as follows: "The affective and schizophrenic forms of mental disorder do not represent the expression of particular pathological processes but rather areas of our personality in which these processes unfold." Thus, schizophrenia and manic depression were not seen as due to particular pathological processes but rather as pre-existing response templates of the human brain to a variety of aetiological factors rooted in genetics and evolution.
In his concept of strata or "registers" of response patterns to pathogenic stimuli, Kraepelin suggested three registers or strata of response:
(1) affective, hysterical and paranoid forms
(2) schizophrenic form
(3) encephalopathic form.
While the affective and schizophrenic forms could easily combine, they would not normally involve demonstrable organic tissue damage. But if a pathological lesion is deep enough to cause an encephalopathic response, it could be expected to activate both the schizophrenic and affective levels of reaction.
Kraepelin's ultimate view of the affective-schizophrenic dichotomy was that "we cannot distinguish satisfactorily between these two illnesses and this brings home the suspicion that our formulation of the problem may be incorrect" (Kraepelin, 1920).
After Kraepelin, the major change in classification of the psychoses came in 1911 with Bleuler, who renamed "dementia praecox" as "schizophrenia". In this new system, certain symptoms were seen as schizophrenic and pathognomonically so—these symptoms defined the splitting of thought from feeling and behaviour; that is, formal thought disorder, blunted affect, autism, and ambivalence. These fundamental symptoms demanded a diagnosis of schizophrenia. Affective symptoms were regarded as non-specific, and manic depression was diagnosed only when schizophrenia was excluded. One of the main effects of the introduction of Schneiderian concepts in the UK in the 1960s was a split in "Anglophone" psychiatry (UK-USA), American psychiatry using a very broad definition of schizophrenia and diagnosing affective patients as schizophrenic.
The convergence of ICD-10 and DSM-IV has bridged the gap. However, as Andreasen (1987) puts it, "The boundary between schizophrenia and affective disorders must remain flexible, depending on whether the goal is research or patient care."
In terms of the conceptualization of bipolar disorders, there has been a discernible split over the boundaries applied to these disorders. Akiskal and others have been calling for a radical extension of the boundaries of bipolar disorder to include various subcategories, such as bipolar II and bipolar III. Klerman (1981) has extended this subdivision by reminding us that mania is not a condition which is the sole preserve of bipolar, but can arise in neurological and toxic states. The following seven subcategories of bipolar disorder are proposed:
type I—mania and depression type II—depression and hypomania type III—mania in response to antidepressants type IV—cyclothymic personality type V—depression with a family history of bipolar disorder type VI—mania without depression type VII—secondary mania.
In contrast, calls have been made to preserve the integrity of the bipolar disorder concept. For example, Baldessarini (2000) criticizes the distinction of the type II bipolar syndrome as lacking any test of its relationship to bipolar I by standard biomedical criteria. Similarly, little has been done in terms of evaluating differential therapeutics.
Furthermore, Baldessarini describes the dilution of the bipolar concept as premature and potentially misleading. He urges restraint, owing to the impression that classical bipolar disorder is "as close to a disease as we have in modern psychiatry. . . . It offers hope of a coherent and tractable phenotypic target for genetic, biological and experimental therapeutic studies."
One of the great debates in nosology has been that of the reality of the difference between bipolar disorder and schizophrenia (Jablensky, 1999). Whether they are two distinct and discrete entities, two partially overlapping clusters of clinical and biological characteristics, or a single continuum is the subject of continuing contemporary debate. The availability of increasingly sophisticated technologies with which to investigate disorders (molecular biology, genetics, and neuroimaging) raises the question of how useful the categorical classification systems, such as ICD-10 and DSM-IV, are in understanding the aetiologies of these disorders.
For example, genome scans of large samples of families with schizophrenic and affective psychoses have identified candidate regions for further study; moreover, several of these regions of interest have loaded for both schizophrenia and bipolar disorder (DeLisi 1999). Similarly, overlaps between the two disorders have been recorded in neuroimag-ing studies (Elliot, 1997). Lastly, epidemiological studies of possible risk factors, such as obstetric/perinatal complications, suggest that these may operate in similar ways across diagnostic/categorical boundaries (Kinney et al., 1993).
These studies represent a clinical-epidemiological approach to the issue of classification. Another approach has been from a neurobiological angle, asking whether, in fact, bipolar disorder and schizophrenia are variations on a theme and represent differing degrees of abnormality within the cognitive-emotional-behavioural circuits that are being mapped currently in neuroscience. Evidence from neuroimaging and from neuropsychology can offer insights into what factors separate these conditions and whether these factors are artefacts of categorization systems, or whether they do represent biological differences both in terms of pathophysiology and in terms of the long-term effects on the brain and thus on prognostication.
In the best circumstances, epidemiological data can be a measure of the distribution of an illness in the population, its extent, and the associated risk factors. Epidemiological data can also link genetic, psychological, environmental, biological, and sociological factors.
There are problems with the epidemiology of bipolar affective disorder, including inconsistencies in diagnosis, treatment, and research design. For example, bipolar disorder is not always included as a separate diagnostic class in epidemiological studies; consequently, the true epidemiological picture remains unclear. However, Goodwin and Jamison (1990) argue that most of the biases in the literature are in the direction of underestimating, rather than overestimating, the incidence and prevalence of bipolar disorder.
Despite methodological variation and consequent interpretative difficulties, a level of agreement is evident. Bipolar disorder is a relatively common condition affecting men and women equally. Cultural, marital, social, and ethnic variation is less clearly defined.
Classification and epidemiology are intimately linked. The accuracy of the latter is essentially dependent on the former. In psychiatry, a branch of medicine which, so far, is devoid of objective physical signs and testing, classification and clinical assessment is all. The epidemiology of bipolar disorder is therefore governed by what classification system is in place and in what way the diagnosis is reached. Modern epidemiology must also take account of the elasticity of diagnostic boundaries, in terms of both broadening and narrowing the definition of bipolar disorder. A summary of factors which influence rates of psychiatric disorders follows:
• breadth of criteria
• instruments used
• lay versus clinical interviewers
• population studied
• single versus repeated observations
• interview of patients versus relatives
• timing of interviews.
While we accept these limitations, establishing the epidemiology of a disorder is an essential part of researching the condition. A crucial factor in this is the completeness of case ascertainment. To achieve complete ascertainment, the options include total population surveys, which are expensive and difficult, especially if the condition under study is rare. An alternative is random sampling, which can have problems surrounding the yield obtained.
Problems more specifically associated with bipolar affective disorder include the problem of "polarity" itself. There is an intrinsic problem in establishing the prevalence of a disorder that can be recognized only at an unpredictable point in its course—namely, when polarity changes.
In their seminal work on bipolar disorder, Goodwin and Jamison (1990) conducted a systematic assessment of the incidence and prevalence of bipolar disorder, using the best studies available to them at that time. The summary statistics from this review showed that the lifetime risk of bipolar disorder was generally less than 1% in industrialized nations (range
0.6-0.9%, 1.2% being a combination of bipolar I and II patients). The annual incidence rate of bipolar was estimated for men at 0.009-0.015%; that is, 9-15 new cases per 100 000 per year. For women, the figures were estimated at 0.007-0.03; that is, 7-30 new cases per 100000 per year.
Perhaps the most influential epidemiological surveys remain the Epidemiologic Catchment Area (ECA) and the National Comorbidity Survey (NCS) (Kessler et al., 1994; Regier et al., 1988). However, future challenges for psychiatric epidemiology include the limitations in currently available surveys as tools for mental health service planning, which include these two large community surveys, the National Institute of Mental Health (NIMH) ECA and the NCS. Both have been the main sources of estimates of treatment need in the USA.
These surveys showed the following prevalence rates of psychiatric and addictive disorders: 1-year prevalence of 30% and a lifetime prevalence of 50%. These very high figures have led to the questioning of their usefulness as proxies for treatment need (Bebbington, 2000; Regier et al., 1998). The high disorder rates were accompanied by low service-use rates, with less than a third of people with active mental disorder using mental health services in a 1-year period (Kessler et al., 1999; Regier et al., 1993). The extent to which untreated cases represent unmet need for treatment, as opposed to absence of the need for treatment because of mild or transient symptoms, is unclear.
Was this article helpful?