THE NOSOLOGICAL STATUS OF SCHIZOPHRENIA: A MULTIDIMENSIONAL VALIDITY STUDY

Abstract

In spite of many years of research, schizophrenia remains a medical puzzle. Even though modern classificatory systems have revolutionized its clinical diagnosis by improving reliability, its validity as a distinct disorder remains controversial. Schizophrenia has many characteristic signs and symptoms but no pathognomonic ones: it has clinical features which overlap those of a number of other disorders. Many workers have been unable to demonstrate a distinct bimodality between schizophrenia and affective psychoses when these disorders have been studied with regard to clinical phenomenology. A number of neurobiologic features have been associated with schizophrenia but most of these are non-specific as they have been demonstrated in patients with other disorders. The course of the iliness is unpredictable and varied. Given this array of unknown, it is no wonder that the Search for its aetiology remains largely exploratory. It is therefore now widely accepted that further significant progress in the study of schizophrenia. especially progress In the identification of aetiological factors and in the possible formulation of preventive measures, will only be made after the problems associated with the clinical heterogeneity of the disorder have been successfully addressed. Over the years, a number of investigators have suggested various ways of subdividing the iliness. These have shown different degrees of success. The challenge of the present time is to use various clinical and neurobiologic domains to determine the most empirically useful way to achieve this subdivision. This study was designed to determine the validity of schizophrenia in a sample of Nigerian patients and to explore which procedure for subdividing it has the potentials for defining broadly homogenous groups of patients. A multidimensional approach was employed such that aspects of the clinical phenomenology, historical and familial antecedents, and neurobiologic correlates of the subjects were Study 60 consecutively admitted patients with schizophrenia diagnosed according to the Research Diagnostic Criteria. Control groups consisted of 53 randomly selected normal subjects, 35 consecutively admitted patients with mania, and 11 with psychotic depression. Study subjects were identical to controls with respect to gender, age, social class, and educational attainments. Patient were made up of non-institutionalized young subject s with relatively short duration of illness. Using standardized procedures, the patient groups were evaluated for a history of perinatal and childhood head trauma, premorbid functioning, psychiatric disorder in first degree relatives, and clinical features of iliness . All the subject had neurobiologic assessment consisting of examination for the presence of neurologic soft signs, determination of cerebral dominance using a battery of sensor motor tests, and evaluation of performance on a number of neuropsychologic tests designed to reflect the functioning of specific brain areas. Assessment procedures anc measurement techniques were shown to be generally reliable. Analyses consisted of comparisons of study and control groups, and of putative subtypes of schizophrenia using univariate statistical methods such as the t-test, chi square, simple correlation, and analysis of variance as appropriate. Multivariate techniques such as various forms of multiple regression analysis were used to determine strengths of associations after potentially confounding factors had been controlled for. Family history data provided support for the view that schizophrenia was a heritable disease. While the first degree relatives of schizophrenics showed an elevated risk for the illness, there was no suggestion that. such relatives showed a similar risk for affective psychoses. Schizophrenics were significantly more likely to have experienced obstetric complications and childhood head injuries than patients with affective psychoses. On the other hand, no evidence was found for a more impaired premorbid functioning, a higher prevalence of neurologic soft signs, or deviant cerebral dominance among schizophrenics than among patients with affective psychoses. Compared with normal and non-schizophrenic controls, patients with schizophrenia showed a relatively consistent pattern of impaired performance on neuropsychological tests, especially on those tests requiring frontal lobe integrity. In a discriminant function analysis, 8 variables consisting of symptomatic features of illness, age at onset of illness, cumulative score of left-sided soft signs, and performance on the Similarities subtest of weschier Adult intelligence test correctly classified 81% of the schizophrenics and 76% of the patients with affective psychoses. Exploration of different subtyping schema revealed that even though the diagnosis of the traditional hebephrenic, catatonic, mixed (or undifferentiated), and paranoid subtypes could be made reliably, these classic subtypes of schizophrenia have little to support their validity in this sample of early onset schizophrenics. There was also no evidence to suggest that subdivisions based on putative aetiolgical factors of familiality or early brain insult have practical utility as only very few patients could be classified in these ways. On the other hand, characterizing patients with the use of positive and negative symptoms profile has empiric validity. However, using factor analytic techniques, a three- rather than a two level typology of syndromes best described the patients. The three syndromes could be termed negative, disorganized, and positive based on their constituent symptom configurations. The results suggested that while the negative syndrome may be characterized by its association with poor premorbid functioning, especially during adolescence, the disorganization syndrome was characterized by its association with widespread impairment of cognitive functioning, particularly of frontal lobe functioning. The positive syndrome generally showed no strong association with the assumed variables. This study has shown that while schizophrenia is a disorder that can be reliably and validly distinguished from normality and for affective psychoses, its clinical heterogeneity often blurs this distinction. Using a multidimensional approach, it was shown that the illness consists of three relatively independent but coexisting syndromes. These schizophrenia were shown to be valid by demonstrating that they bore differential association with historical and neurobiologic correlates. The validity of this approach at subdividing the illness suggests that future research into various aspects of schizophrenia, especially aspects of the aetiology and biological substrate of the illness, are more likely to be rewarding if the disorder is seen as consisting of these three overlapping syndromes.