Background. We sought to develop a clinically useful subtyping system for the non-melancholic depressive disorders, and here we assess one weighted to central aetiological factors.

Methods. We studied 185 patients meeting DSM-III-R and/or clinical criteria for non-melancholic depression. Data were obtained by self-report, interview of patients and from corroborative witnesses. We developed a set of variables for class definition, assessing: (i) ‘P’, disordered personality as a vulnerability factor; (ii) ‘A’, meeting criteria for a lifetime anxiety disorder or positive on probe questions about trait anxiety characteristics, so assessing anxiety as a vulnerability factor; and (iii) ‘L’, psychiatrist and consensually-rated life event stress prior to depression onset.

Results. A latent class analysis generated a four-class solution for the P–A–L variables. Life event stressors had similar item probabilities across all four classes, and did not influence the four-class ‘P–A’ solution when deleted from the analysis, suggesting that life event stress may act more as a general provoking agent, rather than constituting any distinct ‘reactive’ or ‘situational’ depression class. Three classes generated clinically meaningful groupings, reflecting varying contributions of anxiety and disordered personality functioning, and with evidence of differential outcome over the following 12 months.

Conclusions: We suggest that a refined aetiologically-weighted model may assist definition of the non-melancholic depressive disorders, and provide the logic for exploring the comparative utility of differing treatments to identified vulnerability-based classes.

Background. Personality traits have shown considerable heritable components. Association between alleles of a polymorphism in the third exon of the dopamine D4 receptor gene (DRD4) and the personality trait Novelty Seeking has been reported. Recently, in a sample of Swedish non-psychiatric subjects we could not detect any significant relationships between the same polymorphism and Novelty Seeking related scales in the Karolinska Scales of Personality (KSP). However, there was a tendency in the direction of the proposed association. There were also tentative associations between an exon I 13 bp deletion polymorphism and the personality traits Socialization and Guilt.

Methods. We investigated a new Swedish population-based sample (N=167) investigated with the KSP for three DRD4 polymorphisms.

Results. Neither of the previous results were replicated. Combining the previous and the present samples did not give rise to any significant association between DRD4 polymorphisms and personality scales.

Conclusions. The dopamine D4 receptor gene is probably not of importance to the different personality dimensions as measured by the Karolinska Scales of Personality.

Background. There is evidence that high expressed emotion (EE) in relatives of patients with schizophrenia is associated with higher levels of burden of care, and with worse perception of patients' social functioning. However, it is not clear whether changes in EE levels over time are associated with changes in relatives' burden of care and their perception of patients' social functioning.

Methods. Fifty patients with a diagnosis of schizophrenia and 50 relatives were included in the study soon after patients' admission to hospital. Thirty-six relatives and 31 patients were re-assessed 9 months after patients' discharge. Both assessments included patients' symptomatology and relatives' EE levels, burden of care, and perception of patients' social functioning.

Results. Twenty-three relatives (64%) had the same EE level in both assessments, nine (25%) had changed from high to low EE, and four (11%) from low to high EE. Improvement in burden and perception of patients' social role performance were significantly more accentuated among relatives who changed from high to low EE than among relatives who had a stable EE level. Variables that best predicted changes in EE levels were changes in burden scores and number of hours of contact between patients and relatives at follow-up.

Conclusions. Change in EE is associated with change in circumstances and burden. Findings support the idea that EE is better understood in an integrative model.

Background. Personality psychopathology exerts a significant and independent effect on the course of benzodiazepine (BZ) discontinuation, worsening the subjective severity of withdrawal symptoms and significantly increasing the occurrence of early taper failures.

Method. One hundred and seventy-one patients participating in a BZ discontinuation programme were administered several personality measures prior to taper. Patients were stabilized for 3 weeks at their baseline BZ dosage and then tapered off 25% per week over 4 weeks, with the option to extend up to 6 weeks if necessary.

Results. High levels of passivity and dependency as assessed by the MMPI Dependence subcluster, and at a trend level high Eysenck Neuroticism and high TPQ Harm Avoidance contributed significantly to the prediction of benzodiazepine withdrawal severity. Though there was a high correlation between personality measures, psychopathology and adjusted BZ dose, the effects of personality on withdrawal severity was significant, particularly in the initial phases of BZ taper, when taper severity was still relatively mild.

Conclusions. These findings indicate that clinical decisions on how to manage BZ tapering should be guided by personality assessments, in addition to the usual considerations of BZ dosage, residual psychopathology, duration of treatment, etc. The potential for difficulty with discontinuation related to personality traits should be one of the factors weighed in the risk–benefit assessment made in the planning of benzodiazepine treatment for patients with anxious symptomatology.

Background. General practitioners are increasingly involved in the care of patients with long-term psychiatric disorders. We have previously reported that general practitioners are less willing to treat patients with schizophrenia than those without such a diagnosis, but this may have been attributable to a reluctance to treat patients with any psychiatric or chronic illness. We, therefore, examined general practitioners' attitudes to patients with chronic psychiatric or medical illnesses.

Methods. A random sample of 260 local general practitioners were each sent one of our case vignettes which were identical apart from mention of a previous diagnosis of schizophrenia, depression, diabetes or no illness. The general practitioners were asked to indicate their level of agreement with 13 attitudinal statements based on the vignette.

Results. One hundred and sixty-six (66%) of the general practitioners responded to the case vignettes. Those responding to the vignette about the patient with schizophrenia were less happy to have that patient on their practice list and were more concerned about the risk of violence and the child's welfare. Those responding to the depression vignette were more likely to offer the patient antidepressants or counselling; and those who replied to the diabetes case were most likely to refer the patient to a hospital specialist. These differences were not attributable to the personal or practice characteristics of the general practitioners.

Conclusions. Patients with schizophrenia arouse concerns in general practitioners that are not simply due to those patients suffering from a psychiatric or chronic illness. Our results suggest that some patients with schizophrenia may find it difficult to register with a general practitioner and receive the integrated community-based health care service they require. Psychiatrists should provide education and support to general practitioners who look after patients with schizophrenia.

Background. This study used a case–control design to examine the association between unemployment and risk of medically serious suicide attempt.

Method. A sample of 302 individuals who made serious suicide attempts was contrasted with 1028 randomly selected community control subjects.

Results. Individuals who made serious suicide attempts reported higher rates of current unemployment (OR=4·2) than control subjects. This association was similar for males and females. However, even before adjustment for confounding factors it was evident that exposure to unemployment made only a small contribution to suicide attempt risk. The population attributable risk for exposure to unemployment was 7·3%. After adjustment for antecedent childhood, family and educational factors the association between unemployment and risk of serious suicide attempt was reduced but remained significant (OR=2·1), suggesting that common antecedent factors made a large contribution to risks of both unemployment and serious suicide attempt. When both antecedent family and childhood factors, and psychiatric morbidity were taken into account, unemployment was not significantly related to risks of serious suicide attempt.

Conclusion. The results of this study provide support for the contention that much of the association between unemployment and suicidal behaviour is non-causal, and reflects common or correlated factors that contribute to risks of both unemployment and suicidal behaviour. Any remaining association between unemployment and suicide attempt risk appears to arise from the correlation that exists between unemployment and psychiatric disorder.

Background. There still is much debate with regard to the role of psychosocial factors in cancer. Evidence suggests that cancer and coronary heart disease (CHD) may have common causes. This study examined the role of pessimism, anxiety and personality in the development of cancer among men who had been diagnosed with CHD but were free of cancer at baseline.

Methods. Two hundred and forty-six men who were treated for CHD, but were free of cancer, filled out psychological scales at baseline. Patients and their families were contacted after 6–10 years to determine the incidence of cancer.

Results. At follow-up, 12 patients (5%) had been diagnosed with cancer (9 cancer deaths). Development of cancer was unrelated to cardiac pathology but was associated with age [ges ]56 years, poor exercise tolerance, pessimism and anxiety. The rate of cancer was 8/60=13% for men with a distressed personality (type-D) and 4/186=2% for non-type-D men (P=0·002); rate of cancer death was 10% and 2%, respectively (P=0·007). Type-D refers to the interaction between high negative affectivity and high social inhibition. Regression analysis yielded older age (odds ratio 4·6, 95% CI 1·5–14·3) and type-D (odds ratio 7·2, 95% CI 2·9–18·1) as independent prognostic factors for cancer.

Conclusions. Type-D personality was a prognostic factor for the development of cancer in men with established CHD. Psychosomatic research should take a broad enough view of the specific and the global psychosocial variables that may play a role in both cancer and CHD.

Background. Long-term low-dosage dependence on benzodiazepines is traditionally explained by withdrawal symptoms. Previous research has not given much attention to reports that suggest that many patients oppose stopping benzodiazepines long before withdrawal symptoms have developed. This study investigates the scope of and factors associated with this pre-withdrawal treatment insistence.

Methods. Patients receiving long-term low-dosage benzodiazepines in primary care were asked to take a drug-holiday of at least 3 weeks. Sociodemographic, medication, morbidity and attitudinal variables were assessed in addition to the GPs' perceptions of their patients.

Results. Two-thirds of the patients rejected the drug-holiday proposal. Patients who refused a drug-holiday were less educated and were using a higher percentage of long-acting benzodiazepines than patients who accepted the drug-holiday proposal. Those who refused were seen by their GPs as being more complaining, harder to satisfy and less co-operative.

Conclusions. These results provide evidence for drug-seeking or craving behaviour of patients who receive low-dosage benzodiazepine prescriptions. A major problem in benzodiazepine withdrawal occurs before the withdrawal programme has even begun. These data show that benzodiazepine low-dosage dependence should be considered a real form of dependence.

Background. Intrusive memories of stressful events, many involving illness and death, are found in a minority of depressed cancer patients, and may predict the course of anxiety and depression.

Method. Matched samples of mild to moderately depressed and non-depressed cancer patients were followed up after 6 months. Anxiety and depression at follow-up were related to measures of intrusive memories of stressful life events and autobiographical memory functioning that had been assessed at baseline.

Results. Levels of anxiety and depression remained fairly constant over time in the two groups, and the depressed group continued to experience high levels of intrusive memories. The presence of intrusive memories at baseline, and the extent to which these memories were consciously avoided, predicted greater anxiety at follow-up, even after controlling for initial severity of physical and psychiatric symptoms. None of the measures of memory functioning predicted levels of depression at follow-up.

Conclusions. Intrusive memories appear to be a marker of more prolonged psychopathology in cancer patients and may respond to direct therapeutic intervention.

Background. We tested the hypothesis that schizophrenia is primarily a frontostriatal disorder by examining executive function in first-episode patients. Previous studies have shown either equal decrements in many cognitive domains or specific deficits in memory. Such studies have grouped test results or have used few executive measures, thus, possibly losing information. We, therefore, measured a range of executive ability with tests known to be sensitive to frontal lobe function.

Methods. Thirty first-episode schizophrenic patients and 30 normal volunteers, matched for age and NART IQ, were tested on computerized test of planning, spatial working memory and attentional set shifting from the Cambridge Automated Neuropsychological Test Battery. Computerized and traditional tests of memory were also administered for comparison.

Results. Patients were worse on all tests but the profile was non-uniform. A componential analysis indicated that the patients were characterized by a poor ability to think ahead and organize responses but an intact ability to switch attention and inhibit prepotent responses. Patients also demonstrated poor memory, especially for free recall of a story and associate learning of unrelated word pairs.

Conclusions. In contradistinction to previous studies, schizophrenic patients do have profound executive impairments at the beginning of the illness. However, these concern planning and strategy use rather than attentional set shifting, which is generally unimpaired. Previous findings in more chronic patients, of severe attentional set shifting impairment, suggest that executive cognitive deficits are progressive during the course of schizophrenia. The finding of severe mnemonic impairment at first episode suggests that cognitive deficits are not restricted to one cognitive domain.

Background. Despite the increasing importance of quality of life in the mental health field, the theoretical conceptualization of the construct remains poorly developed. A proposed mediational model of quality of life, which links subjective quality of life with self-related constructs, is examined with a group of long-term psychiatric hostel residents. The present study aims to develop a measure of quality of life based on the proposed model, to explore the data and their implications for service development and finally to conduct a preliminary analysis of the model's predictions.

Method. A cross-sectional research design was employed. Quality of life interviews, using a modified version of Lehman's Quality of Life Interview, were carried out with 54 psychiatric residents in Greece. The model's predictions were examined by using a series of regression analyses.

Results. The results indicate that perceived improvements in lifestyle, greater autonomy and positive self-concept are significantly and directly associated with better quality of life. In contrast, a direct relationship between objective indicators and subjective quality of life was not found.

Conclusions. The traditional two-part quality of life model that includes objective indicators of life circumstances and subjective indicators is extended to included the constructs of self-concept and perceived autonomy. The present extended mediational model of quality of life for individuals with long-term mental health problems appears to have important implications for the planning and delivery of mental health programmes.

Background. Recent anthropological studies have documented the importance of understanding the relation of culture to the experience of mental illness. The use of interviews that elicit explanatory models has facilitated such research, but currently available interviews are lengthy and impractical for epidemiological studies. This paper is a preliminary report on the development of a brief instrument to elicit explanatory models for use in field work.

Method. The development of the SEMI, a short interview to elicit explanatory models is described. The interview explores the subject's cultural background, nature of presenting problem, help-seeking behaviour, interaction with physician/healer and beliefs related to mental illness.

Results. The SEMI was employed to study the explanatory models of subjects with common mental disorders among Whites, African-Caribbean and Asians living in London and was also used in Harare, Zimbabwe. Data from its use in four different ethnic groups is presented with the aim of demonstrating its capacity to show up differences in these varied settings.

Conclusions. The simplicity and brevity of the SEMI allow for its use in field studies in different cultures, data can be used to provide variables for use in quantitative analysis and provide qualitative descriptions.

Background. A Dutch cohort of predominantly out-patient DSM-III-R major depressive patients was followed for 3 to 5 years after start of treatment in a psycho-neuro-endocrinological prediction study. The study design permitted description of the course of remissions, relapses and recurrences.

Methods. Pharmacological treatment was standardized, psychotherapy was tailored to the needs of the patient, follow-ups were done monthly until 3 years or more after the initial recruitment.

Results. After 9 months 49% of the patients had reached full remission and 45% were in partial remission. During the following 3 to 5 years 82% of the patients had reached a period of full remission. Sixteen per cent of the patients needed 2 years or more before full remission. A relapse or recurrence rate of 41% within 5 years was found. Patients with residual symptoms relapsed particularly in the first 4 months after remission, while patients without residual symptoms recurred mainly after 12 months after remission. Previous depressive episodes and psychoticism predicted relapse. Psychomotor retardation at inception predicted a longer time to partial remission.

Conclusion. In most cases, major depression is a seriously impairing episodic disease. This is also true for a sample of predominantly out-patients treated at a university clinic.

Background. Causal attributions about bodily sensations may determine help-seeking and influence patients' demands for medical treatment. The present study aimed to differentiate the causal attributions associated with health-related and non-health-related anxiety.

Methods. Anxious hypochondriacal, generally anxious, and non-anxious general practice attenders were compared on their propensity to give somatic, psychological or normalizing attributions for common bodily sensations, measured by number of each type of attribution in a given time period and the frequency of first response of each type.

Results. The groups differed in all three types of attributions. Giving more psychological and fewer normalizing attributions was related to general anxiety whereas giving more somatic attributions was related specifically to hypochondriasis.

Conclusions. Anxiety and hypochondriasis can be distinguished in terms of their associated patterns of attributions for bodily sensations, reinforcing the importance of attributional processes and interventions which use reattributional training.

Background. There is increasing awareness of the prognostic value of residual symptomatology in affective disorders and of the need for specific therapeutic strategies in this phase of illness. The aims of the study were to apply a novel, short-term psychotherapeutic approach for increasing well-being, based on Ryff's conceptual model, to remitted patients with affective disorders and to compare the results with those obtained with symptom-oriented cognitive behavioural strategies.

Methods. Twenty patients with affective disorders (major depression, panic disorder with agoraphobia, social phobia, generalized anxiety disorder, obsessive–compulsive disorder) who had been successfully treated by behavioural or pharmacological methods were randomly assigned to a well-being enhancing therapeutic strategy (well-being therapy) or cognitive–behavioural treatment of residual symptoms.

Results. Both well-being and cognitive–behavioural therapies were associated with a significant reduction of residual symptoms. However, a significant advantage of well-being therapy over cognitive–behavioural strategies was observed with observer-rated methods.

Discussion. These preliminary results suggest the feasibility of well-being therapy in the residual stage of affective disorders. Further research should determine its value as a relapse-preventive strategy in specific mood and anxiety disorders.

Background. The aim of the current study was to evaluate the utility of maternal recall for obtaining history of obstetric complications (OCs) in psychiatric research.

Methods. Obstetric information from in-depth structured maternal interviews and from hospital birth records was compared in 45 mothers of schizophrenic patients and 34 control mothers. Both types of information were blindly and independently scored for OCs using the McNeil-Sjöström OC Scale.

Results. Considerable discrepancies were observed between interviews and records, irrespective of maternal group. No significant differences were found between patient and control mothers in error type (omission, commission, total retrospective error) or recall facility for selected events. More errors of commission were made by mothers of patients with negative family history than mothers of patients with positive family history for psychiatric disorder. Patients had significantly more OCs than controls only when hospital record information was utilized.

Conclusions. OC history obtained by maternal recall has methodological limitations.

Background. We examined interleukin-2 (IL-2) related immune pathways in depression to elucidate mechanisms underlying various immunological disturbances associated with depression.

Methods. Subjects comprised 35 unmedicated patients with a major depressive episode without psychotic features and 36 age- and sex-matched healthy volunteers. The immune parameters examined included the numbers of B and T cells, IL-2 receptor-mediated blastoformation (IL-2R-mediated blastoformation), IL-2 production and expression of the IL-2 receptor α-subunit.

Results. The patients with a severe episode showed significantly lower IL-2R-mediated blastoformation than the controls. There was a statistically significant negative correlation between IL-2R-mediated blastoformation and the severity of depression at the time of entry.

Conclusion. The reduced IL-2R-mediated blastoformation may partly explain several previously reported abnormal immune functions associated with depression.