Since descriptive psychiatry offers only the surface exploration of psychopathology we must turn to aetiology for the deeper understanding of our problems. It has become quite clear during the last few decades that the factors in the aetiology of schizophrenia are less established than those in the efficacy of therapeutic intervention. The former are still highly theoretical scientific models whose tenability is often difficult to establish, while the latter can be tested more objectively through clinical trials. In this manner, the efficacy of neuroleptic therapy and of the various behavioural and psychosocial therapeutic interventions has been examined and some found to be wanting (e.g. renal dialysis). Nevertheless, the therapeutic clinical trials probably arose as a result of some aetiological considerations and for new therapeutic approaches we need to turn again to our aetiological models. However, the connection between aetiological models and therapeutic intervention has never been examined. As a result we have elaborate aetiology on the one hand and specific therapies on the other but the two pass each other by. The purpose of this paper is to weave the threads of practice into the web of theory.

The scope of this introductory contribution is to present some general facts and ideas which could enhance our understanding of this enigmatic illness. By ‘mediating processes’ (or ‘mediators’) I mean mechanisms providing bridges between different aspects of knowledge of the schizophrenia-puzzle which, so far, have not been linked together in a satisfactory way: one of the central problems is the identification of mediators between psychosocial factors on the one hand, and biological factors on the other. If we do not succeed in building up conceptually valid bridges between these two aspects, which are testable for hypothesis-orientated research, a disastrous splitting of our understanding and treatment of schizophrenia into either biological or psychosocial reductionism is likely.

To paraphrase Kant, theory without data is mere intellectual play, but data without theory are blind. “Mediating Processes in Schizophrenia” is a topic demanding theory as well as data. Although the topic may seem limited in scope, it actually probes the causes of schizophrenia. The causes involved in ‘mediating processes’, however, are not the usually considered primary or distal causes of schizophrenia, such as genetics or early environment. Instead they are the more proximal causes – mechanisms that are closer to illness onset, exacerbation, or improvement. It is amazing how little attention is paid to this kind of cause.

In a study of children adopted at an early age, discrimination between hereditary and family dynamic factors is possible. The biological parents have given the child their genetic characteristics and sometimes the very early environment, while the adoptive parents have provided the more permanent family environment and rearing. The major goal of the Finnish adoptive family study was to re-assess the genetic contributions to schizophrenia and to provide further measures of the adoptive family rearing environment. In other words, we were interested in the joint effects of genetic and family environmental variables and in their possible contribution to both the psychopathology and the healthy functioning of adoptees during their development. This approach paid attention to the possibility that a healthy, possibly protective, rearing family environment may reduce the genetic risk. During the course of the study, we also considered whether the direction of the effects between genetic and family environmental factors can be clarified through a prospective, longitudinal study of the adoptees at risk.

The pharmacotherapy of psychotic disorders which started in 1952, based on empirical results obtained with chlorpromazine by Delay & Deniker, has revolutionised the treatment of psychiatric diseases of non-organic origin. The discovery of haloperidol by Janssen in 1959 represented further progress in this field, since this was found to be more potent and to have generally less disturbing adverse effects than chlorpromazine. Nevertheless it took several years before dopamine-antagonism was put forward as possible mechanism of action for neuroleptic drugs (Carlsson & Lindqvist, 1963). From the early 1970s, however, the advent of radioligand binding studies provided new tools for studying neurotransmitter receptors in the brain and for investigating the interaction of drugs with various receptors. This technique allowed identification of receptor-binding sites related to pharmacologically defined receptors and receptor subtypes, e.g. α and β-adrenergic receptors. It also led to further subclassification and refined definition of receptors, e.g. dopamine-D1 and -D2, and different pharmacological effects mediated by the receptor subtypes have been identified.

Hypotheses as to the pathogenesis of schizophrenia can be discussed at different levels of a possible manifestation of the causative factor: the macroscopic-morphological, the microscopic-morphological, and the molecular. Some abnormalities have been observed on all of them: e.g. increased ventricular-brain ratios in CT, hypofrontality in SPECT and in glucographic PET-scans, and other macromorphological abnormalities (for reviews cf. Bogerts 1984; Mundt, 1986; Bogerts et al, 1987), gliosis on a microscopic level (Stevens, 1982), and an increased dopamine-binding in in vivo receptor studies (PET as well as in post-mortem studies; Cazzullo, 1988). However, the diversity and variability of these findings point to the view that rather than there being a single distinct pathogenetic factor responsible for the pathogenesis of schizophrenic psychoses, a constitutional disposition exists, which can be described as a functional dysequilibrium within the human brain. From this point of view, schizophrenia would not appear as an inherited disorder of metabolism, but as a weakness of a neurobiological ‘system’, i.e. as an interactional disorder of a complex of networks, in which the interaction between different substructures is labile in such a way that under special conditions (e.g. ‘stress’), a decompensation (functional breakdown) results. In this sense, ‘vulnerability’ to schizophrenia may be interpreted as a consequence of a constitutional deficiency of the brain which results in an inability to stabilise, under specially challenging conditions, the interaction between different substructures of the human brain. Before this ‘functional dysequilibrium-hypothesis’ (which is a special form of a constitutional structural deficiency-hypothesis) is discussed, and before the question is raised as to which are the relevant dysequilibrated components, some indication will be given as to why such an hypothesis appears plausible.

Current interest in vulnerability to schizophrenia has been stimulated mainly by the work of Zubin and associates (Zubin & Spring, 1977; Zubin & Steinhauer, 1981). Figure 1, which was first published by Zubin & Spring in 1977, illustrates the basic idea common to every diathesis-stress model: The outbreak of a psychotic episode cannot be explained solely by referring to environmental events: a susceptibility of the individual has also to be taken into account. Ways to identify vulnerable individuals were discussed by Zubin & Steinhauer (1981), who introduced the important distinction between trait as opposed to episode markers. Only stable characteristics relate to vulnerability.

The presence of non-psychotic neurotic and dysphoric symptoms in schizophrenia has been recognised from the days of Bleuler (1908) and Kraepelin (1896) and is inherent in the hierarchical schema explicated by Foulds (1976) and commonly employed by psychiatrists in the diagnosis of psychotic disorder. Of such symptoms, those which may be characterised as depressive in nature have been the focus of much recent interest in view of claims that they may be iatrogenic. That such symptoms occur widely in schizophrenia is beyond dispute, but their relationship to the process of the illness and its treatment remains a subject of contention.

This paper attempts to describe some of our electrodermal data and its relationship with environmental factors and the outcome of a psychosocial intervention study with families of schizophrenics. In the context of the heuristic vulnerability/stress model of the development and course of schizophrenic episodes described by Nuechterlein & Dawson in 1984, there is an interaction between individual vulnerability factors and external environmental stressors with resulting effects on the hypothesised transient intermediate states of: processing capacity overload, autonomic hyperarousal and a deficient processing of social stimuli. Accelerating dysfunctions in these systems result in the reappearance of characteristic schizophrenic psychotic symptoms. The purpose of our work has been to investigate the interface between the individuals' biological and environmental systems.

Around the turn of the last century, Kraepelin and Bleuler developed the theory that some schizophrenic symptoms can be traced back to attentional disorders and in contemporary experimental research on attention, this reappears in the thesis that attentional disorders cause a vulnerability to schizophrenia. Empirical studies (Nuechterlein & Dawson, 1984) showed that attentional deficits before, during, and after an episode of such illness are consistent with the vulnerability model. However, if attentional disorders really do play a fundamental role in the development of schizophrenic symptoms, how do patients handle their more-or-less intact attentional capabilities? Thus, at the current level of research, the relevant question is not only whether attentional disorders occur, but – more and more – how such patients process information in the presence of these disorders.

Cognitive deficits in schizophrenia can be differentiated into state-linked accompaniments of psychotic episodes and trait-linked signs of a more enduring vulnerability to schizophrenia (Zubin & Spring, 1977; Cromwell & Spaulding, 1978). Elsewhere (Spring & Zubin, 1978) we suggested that a marker of psychotic episodes is one that deviates in highly symptomatic schizophrenic patients and normalises as their florid psychotic symptoms remit. A stable vulnerability marker, in contrast, continues to deviate in remitted patients, their biological relatives, and others who are vulnerable to schizophrenia but not floridly psychotic. In addition, Nuechterlein & Dawson (1984) proposed a mediating vulnerability marker which demonstrates characteristics of both vulnerability and episode markers. On such a marker, asymptomatic vulnerable individuals differ from normal; the advent of psychosis magnifies the deviation.

This discussion follows from a key premise that there is potential clinical value in identifying and directly treating the cognitive abnormalities which for over a century have been seen as the hallmarks of schizophrenia. This is both an old and a relatively new idea. It is old, in that the psychodynamic approaches to psychotherapy, especially ego psychology, have long sought to modify cognition in schizophrenia, if unsystematically and indirectly. It is new, in that only in the last decade have both the contemporary technologies of the information processing laboratory and of cognitive-behaviour modification been applied to the problem of treatment. The failures of psychodynamic treatment approaches with schizophrenic patients have perhaps discouraged widespread experimentation with cognitive models in clinical assessment and treatment. This is unfortunate, as contemporary clinical procedures and laboratory technologies are hardly comparable to psychodynamic methods. In the past decade findings have accumulated, in the form of case reports, single-subject experiments and treatment trials, which support the potential usefulness of assessing and directly treating schizophrenic patients' cognitive abnormalities (reviewed by Spaulding et al, 1986).

Schizophrenia research is affected by the current change of scientific paradigms of thinking and investigation. The 1960s and 1970s were mainly characterised by a theoretical shift from a unifactorial towards a multifactorial view of the aetiology of schizophrenia, while in the 1980s, traditional linear models of pathogenesis have been more and more given up in favour of systems models.

The general view that relationships between factors within the patient and those in the environment are important in understanding schizophrenia has been widely accepted and is the basis for vulnerability/stress models of this disorder. However, the specifics of the patient-environment relationships that may be critical for schizophrenia continue to be largely a mystery. At the UCLA Clinical Research Center for the Study of Schizophrenia, we have developed a tentative working model of schizophrenic episodes that emphasises the mediating role of information-processing and autonomic abnormalities in the patient in interaction with stressful circumstances and protective factors in the patient's environment (Nuechterlein & Dawson, 1984a; Liberman, 1986; Dawson & Nuechterlein, 1987; Nuechterlein, 1987). Here we examine two examples of patient-environment relationships that may increase our understanding of such processes in schizophrenia.

Attention was first drawn to the importance of this topic by Schneider (1957) when he proposed that some of his first-rank symptoms could be grouped together under the concept of ‘permeability’ of the barrier between the individual and his/her environment — the ‘loss of ego boundaries’. Similarly, in a recent series of papers (Strauss et al, 1987), the essence of schizophrenia was conceptualised in the processes of interaction between biology, behaviour, and environment. The two main syndromes — acute and chronic — can each be precipitated or made worse by environmental factors, and although most patients with the negative syndrome appear to show some irreducible impairment, poverty of the social environment has been found to worsen their condition, while on the other hand, a moderate degree of social stimulation promotes relative improvement (Wing, 1987).

The research question of which mediating factors influence the long-term course of schizophrenia was not asked until recently because the expectation has been of uniformly poor outcome (Kraepelin, 1902; American Psychiatric Association, 1980). However, anecdotal clinical knowledge about heterogeneity in the long-term course of this severe illness has been firmly supported in the last 15 years by six recent longitudinal studies in Europe, Asia, and the USA (Blueler, 1972; Ciompi & Müller, 1976; Huber et al, 1979; Tsuang et al, 1979; Harding et al, 19876; Ogawa et al, 1987). Each of these studies, more methodologically rigorous than those of the past, has found multiple indices of wide heterogeneity in the long-term outcome of schizophrenia with trends toward significant improvement or recovery demonstrated in over half of each cohort.

A number of recent studies (Leff & Vaughn, 1985; Nuechterlein et al, 1986; Jenkins et al, 1986; Mintz et al, 1987) have indicated that attitudes of family members towards a relative with schizophrenia (termed High Epressed Emotion (EE)), are predictive of the short-term course of the patient's disorder. The measure used to define EE status of a relative is a semi-structured interview, the Camberwell Family Interview (CFI) (Vaughn & Leff, 1976), which is usually administered shortly after the patient is admitted to hospital for an index episode of the disorder.

Hypotheses on the relationship of schizophrenia and family variables have changed considerably over the last 15 years: whereas speculations on the causal role of familial interaction for the onset of schizophrenic psychosis previously dominated the field of psychological theorising and psychotherapy (Bateson et al, 1956), it was not possible to confirm these theories empirically. In accordance with the research on Expressed Emotion (EE), a shift in emphasis to the influence of family variables on the further course of the illness has taken place. As a consequence, promising new techniques have been developed for the prevention or postponement of relapse.