Present drug treatments for the more common types of dementia are largely palliative or symptomatic, but the promise of drugs to prevent the development of dementia does not seem unrealistic. Neuropharmacological strategies that aim to improve behavioural or cognitive symptoms in dementia are subdivided into drugs that modify central cholinergic, serotonergic, dopaminergic, GABAnergic, and peptidergic transmission. There are no clear clinical guidelines for the management of demented patients who are also hypertensive. The early identification of patients with incipient dementia, the detection of slight but theoretically important beneficial effects, the optimum method of drug administration, more soundly based treatment of behavioural symptoms, and the evaluation of drug combinations are all identified as areas requiring further research.

Existing antidepressants are unsatisfactory: they have limited efficacy, side-effects are a problem, and there is an appreciable degree of toxicity. However, nonpharmacological, psychological approaches to the treatment of depression also have their own drawbacks, so that the search for better antidepressant compounds continues.

Although suicide was traditionally considered an extreme response to stress, with the most frequent stress being depressive illness, severity of depression does not distinguish those who commit suicide from non-suicide attempters. A biological role involving the serotonergic system, possibly associated with a genetic risk factor, has been postulated. Low levels of 5-HT and 5-HIAA have been found in post-mortem examinations of brain-stem tissues of suicide victims (levels in cortical tissue were generally normal). An increased number of 5-HT2 receptors was found in the pre-frontal cortex of suicide victims, such upregulation having been demonstrated in induced 5-HT deficiency states; 5-HT1A receptors were also increased. Receptor populations may be altered by chronic psychotropic medication; e.g. 5-HT2 downregulation occurs following administration of antidepressants. There is some indication that the adrenergic system may be involved as well, but this will require further study. Antidepressant drugs may be effective in preventing suicide in patients with non-depressive syndromes who exhibit suicidal behaviour.

The dopamine hypothesis still provides a valuable approach to the study of schizophrenia and its treatment by drugs. Although the neuroleptic drugs appear to act via an inhibition of dopamine receptors, measurements of dopamine metabolites in vivo, or of the transmitter and its receptors in postmortem brain tissue, do not provide unequivocal evidence of a hyperactivity of dopaminergic neurotransmission in the disease. Nevertheless, increased dopamine function might be a consequence of a primary neuronal abnormality in another system. Recent imaging studies and neuropathological reports suggest that, in some patients, there may be a deficit and/or disturbance of neurons in certain temporal limbic regions, and this is supported by some neurochemical investigations, particularly of neuropeptide and amino-acid transmitter systems. A loss of such neurons could conceivably lead to a disinhibition of limbic dopamine neurons, providing the means whereby neuroleptic drug treatment might ameliorate the effects of a neuronal deficit in schizophrenia.

The psychopathology and disabilities of a group of institutionalised elderly schizophrenics were compared with those of a group of non-schizophrenic elderly, and with younger schizophrenics reported in a parallel study and in the literature. Elderly schizophrenics showed significantly more psychopathology and disability than the non-schizophrenic elderly; however, they exhibited less psychopathology than younger schizophrenics. Disability among elderly schizophrenics was not correlated with chronicity of institutionalisation or psychopathology. There was a suggestion that disability among elderly schizophrenics may be due to organic deterioration.

In a recent edition of the Schizophrenia Bulletin devoted to negative symptoms, the editor suggested at the beginning of his preface that what he called the “positive–negative symptom distinction” had been introduced into psychiatry only a decade previously (Levine, 1985). Many authors in the same volume seemed to agree with him; at least they did not emphasise any continuity with earlier work. There were a few notable exceptions. Since, for more than 30 years, I have seen the relationship between psychological deficit (cognitive defect, negative syndrome) and the productive (florid, positive) symptoms as lying at the heart of the mystery of schizophrenia, it seemed possible that some account of how that interest arose and developed might still be timely. This is part of the history of the Social Psychiatry Research Unit of the Medical Research Council.

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.

Evidence is provided in support of the nosological concept which distinguishes schizophrenia from borderline states. Differentiation of these disorders by a computer pattern-recognition program was based on the initial descriptive characteristics of patients. As a result, we observed good correlation between computer diagnosis from data on 215 patients and follow-up data analysed according to the criteria of Kraepelin's model of schizophrenia and its course. According to the findings on outcome, the states designated as 'sluggish schizophrenia’ (or 'slowly developing variant of schizophrenia’) comprise a group of dissimilar psychiatric disorders, including borderline states and cases of process schizophrenia, with the specific initial symptoms of the latter disease.

Volunteer college students completed a categorisation task designed to test the hypothesis that adults organise their expectations of relationships in a manner that reflects a functional distinction between attachment and affiliation. Terms associated with the provision of security and with characteristics theoretically identified as definitive of attachment relationships were significantly more likely to be associated with the relationship identified as ‘lover’. There was substantial overlap between attachment and affiliation relationships in functions relating to intimacy and disclosure. These results highlight the need to differentiate close relationships by functional rather than structural characteristics.

My aim in this lecture is to review a wide range of published studies on the treatment of depression. I want to look at these with one major clinical question in mind: given the range of treatments available, which treatment should be chosen for which patient? In July 1985 a report, published as a special supplement to the American Journal of Psychiatry, called for more research on this topic:

“As the number and diversity of the effective drugs and psychosocial interventions expand, clinicians will increasingly want objective, reliable criteria with which to match the patient to the optimal treatment modality and to monitor the response.” (Institute of Medicine, 1985)

A sample of 890 people aged 65 years and over living at home in the Lewisham and North Southwark Health District was interviewed using a structured schedule containing CARE organic brain syndrome and depression scales, and items dealing with anxiety symptoms. Total prevalence rates were 4.6% for cognitive impairment, 13.5% for depression, 3.7% for generalised anxiety, and 10.0% for phobic disorders. Cognitive impairment increased exponentially with age, and was associated with lower occupational class. There was no significant association between occupational class and depression, generalised anxiety, or phobic disorder in this sample. Cognitive impairment, depression and agoraphobia were associated with significantly higher levels of dependency in the tasks of daily living.

“Health is much more difficult to deal with than disease.” D. W. Winnicott

I

In a talk given in 1945 to the sixth form of St Paul's School, Donald Winnicott described his experience, as a schoolboy, of discovering Darwin's Origin of Species:

The development of an investigator-based standardised interview to assess patterns of specific and general social dysfunction is described. It covers six domains of functioning: work; love relationships; friendships; non-intimate social contacts; negotiations; and everyday coping. Inter-rater reliability was tested by three investigators rating 21 audiotaped interviews, and was shown to be high, with an intraclass correlation of 0.87 for the total score. The pattern of associations between specific and general social dysfunction was examined through determination of sensitivities and specificities and through LISREL modelling. The findings varied across social domains but it was concluded that the total APFA score provided a reasonable measure of general social dysfunction.

A total of 1269 psychiatrists, from 15 European countries, completed a questionnaire assessing treatment choice in depressive and anxiety disorders, with particular focus on diagnostic and other factors determining use of monoamine oxidase inhibitors (MAOIs). Tricyclic and similar antidepressants were used most commonly — MAOIs were used as first-choice therapy for depressive or anxiety disorders only by a minority but were commonly a second-choice treatment for atypical depressives. Use was inhibited by interactions and side-effects rather than by perceived ineffectiveness. Cross-national differences in treatment were considerable but diagnostic differences in case history vignettes were much smaller.

Recent interest in positive and negative symptoms in schizophrenia is fuelled by the hope that this dichotomy will identify components with differing prognostic, therapeutic or pathogenetic significance. Correlations have been sought, and some found, in each of these areas. Yet the origin and definition of negative symptoms remain a matter of debate.

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.

General hospital psychiatry in Britain began in 1728, and thereafter several new voluntary hospitals provided separate wards for lunatics, but none survived beyond the middle of the 19th century. Less severe nervous organic disorder has always been common in the general wards of voluntary hospitals, and was accepted as the responsibility of neurologists and other physicians; all forms of disorder were admitted to the infirmaries of workhouses. During the present century psychiatrists began to take an interest in non-certifiable mental illnesses and in working in general hospitals. Out-patient clinics became more common following the Mental Treatment Act 1930. The growth of general hospital psychiatric units in the last 30 years began amidst controversy, but has received little recent critical attention.

We have attempted to render this article into intelligible English. The original was generally somewhat cumbersome. On occasion, a more polished translation might have been attempted, but as the reader should be able to estimate the author's intent, we have avoided any possible liberties in interpretation.

Plasma concentrations of prolactin, growth hormone, cortisol, TSH, and the neurophysins were measured over 17 hours in 98 newly admitted psychiatric patients and 35 control subjects. Seventy patients had been free of psychotropic medication for three months. Patients with schizoaffective mania (SAM) differed significantly from control subjects by increased plasma Cortisol concentrations and decreased night-time TSH concentrations. The latter were also significantly lower than in both schizophrenic and manic disorder patients. Plasma Cortisol was increased to a lesser extent in other psychotic subgroups, and increases in prolactin were most marked in the affective psychoses. There was little diagnostic specificity for psychoses other than SAM. Higher Cortisol and prolactin levels may be due to the stimulatory effect of serotonergic pathways, but the neural mechanisms underlying lower night-time TSH levels in SAM are not known. The findings are not consistent with the view (a) that the hormonal changes of the psychoses simply reflect a non-specific response to stress, or (b) that the biological abnormalities of the psychoses can be accounted for by a single continuum of disturbance.

Involvement of the brain serotonin (5-HT) neurotransmitter system in obsessive-compulsive disorder (OCD) was originally suggested on the basis of therapeutic effects found with the semi-selective serotonin uptake inhibitor, clomipramine. More recent studies directly comparing clomipramine with non-selective or norepinephrine-selective uptake inhibitors, such as desipramine or nortriptyline, as well as studies with new, more selective serotonin uptake inhibitors, including fluvoxamine and fluoxetine, have supported that hypothesis. Clomipramine's antiobsessional effect has been augmented with the serotonin precursor, L-tryptophan, or with lithium, which has prominent serotonergic effects. Patients whose OCD symptoms improved on clomipramine worsened when the drug was discontinued (regardless of duration of therapy) and improved when clomipramine was reinstituted. OCD symptoms also worsened when metergoline, a 5-HT antagonist, was given to patients who had improved with clomipramine. Metergoline given alone had no effect. Administration of m-chlorophenylpiperazine (m-CPP), a 5-HT receptor agonist, to untreated OCD patients increased their anxiety, depression, and dysphoria, and exacerbated their OC symptoms. After 4 months of clomipramine therapy, m-CPP failed to produce the same behavioural effects, suggesting an alteration of a 5-HT subsystem (possibly downregulation of some 5-HT receptors). The data reviewed suggest an important role for an abnormal brain 5-HT subsystem in patients with OCD.