The first issue of 1997 brings two changes.

First, we are pleased to announce a further increase in size. Pressure of excellent quality manuscript submissions continues to grow, and this year sees a modest increase in size with each issue around 256 pages. With an earlier increase from four to six issues per year, this represents 40% growth in the total size of each volume in the last 3 years. We recognize the importance of keeping publication delays short, and also the importance of maintaining high quality. The new increase in pages will allow only a modest increase in our current acceptance rate, without loss of these ideals, and the annual volume size is now substantial.

Secondly, we adopt structured abstracts with this issue. This format has become widespread in the last few years. It is better suited to the electronic databases, which are essential for the modern scientist and scholar, and to the various forms of electronic publishing, which are supplementing, but not supplanting, the printed page. We are proud of the literary style of our authors and we hope that the structured format can still accommodate it.

‘Whether or not it is worthwhile to distinguish between “neurasthenia” and “dysthymic disorders” must depend either on the demonstration that such syndromes have different social covariates, or pursue a different course, or have particular responses to treatment. Until such studies are forthcoming, the distinction seems especially insubstantial.’ (Goldberg & Bridges, 1991)

Epidemiological studies now indicate that fatigue is one of the most common symptoms of ill-health in the community, primary care and other medical settings, and that syndromal diagnoses based on fatigue (including chronic fatigue and neurasthenia) are prevalent and major sources of health care utilization. Such syndromes are characterized by a combination of persistent and disabling fatigue and neuropsychological and neuromuscular symptoms (Lloyd et al. 1990; Angst & Koch, 1991; Sharpe et al. 1991; Fukuda et al. 1994). Essentially, the differences between these syndromes reflect variations in duration criteria rather than symptom constructs. Specifically, the Centers for Disease Control (CDC) defines ‘prolonged fatigue’ as a syndrome of at least 1 month's duration, and chronic fatigue (including idiopathic and chronic fatigue syndrome – CFS) as a fatigue syndrome of at least 6 months duration (Fukuda et al. 1994). The ICD-10 classification system (World Health Organization, 1992) now includes a formal diagnosis of neurasthenia (F48.0) based on mental and physical fatigue of at least 3 months duration. Despite the current international and epidemiological interest in these disorders, DSM-IV has simply included them within the ‘Undifferentiated Somatoform Disorders – 300.81’ category (American Psychiatric Association, 1994).

There are strong moral overtones associated with eating behaviour and with this goes the contemporary assumption that anyone can choose their shape and weight if only they try hard enough. Arguably, it is this false gospel that has led to the recent increase in bulimia nervosa (BN). Women struggling to attain an idealized shape have drastically to subvert their bodies' homeostatic mechanisms: the battle is rarely won and eventually, the victim turns for help to escape the scars and trauma. On the other hand, the shimmering vision of anorexia nervosa (AN) appears to support the myth.

Non-invasive physiological measures of in vivo brain function, derived from positron emission tomography (PET) and functional magnetic resonance (fMRI), are now standard tools in cognitive neuroscience. These techniques provide a powerful context for addressing critical questions regarding both the localization and mechanisms of higher brain functions. A general overview of the history and development of imaging techniques in relation to cognitive science is that of Posner & Raichle (1994). The conceptual background and methodological approaches available for neurobiological based psychiatric research, in conjunction with functional imaging, provides the focus for the present review.

Major psychiatric disorders, such as schizophrenia and depression, represent disturbances at the highest level of brain function. Providing a neurobiological account of these conditions presents the most formidable problem in clinical neuroscience. Questions posed by neurobiological perspectives on psychiatric disorders are necessarily embedded in theoretical assumptions about how we think that the brain works. From this, it follows that an important limiting factor in any neurobiological account of psychiatric disease is the stage of development and theoretical conceptualization of brain function in general.

From a public health perspective, there is probably no more important or daunting challenge than reducing the prevalence of the most common mental disorders, namely anxiety and depression. These disorders have a combined community prevalence rate of between 15% and 30% (Cox et al. 1987; Robins et al. 1991; Goldberg & Huxley, 1992; Kessler et al. 1994; Meltzer et al. 1995) and account for one-third of days lost from work due to ill health (Jenkins, 1985a) and one-fifth of all consultations in general practice in the UK (Williams et al. 1986). The common mental disorders are associated with impairments in physical and social functioning at least as severe those associated with chronic physical illnesses (Wells et al. 1988; Klerman, 1989; Wohlfarth et al. 1993; Ormel & Costa e Silva, 1995; Spitzer et al. 1995; Martin et al. 1996), and a mortality rate nearly twice that of the general population (Murphy et al., 1987; Klerman, 1989; Lloyd et al. 1996). The public health importance of these disorders, even in mild form, is further demonstrated by the finding that low levels of depression resulted in 51% more days lost from work than major depression (Broadhead et al. 1990). The total annual cost of the common mental disorders in the UK may amount to £6 billion, of which two-thirds arises from lost productivity (Croft-Jefferys & Wilkinson, 1989). Despite the availability of simple, cheap and effective treatments (Paykel & Priest, 1992; Effective Health Care, 1993; Brugha, 1995), one study found that the prevalence of these disorders in Britain may have increased recently (Lewis & Wilkinson, 1993).

Depression has a profound impact on parameters of interpersonal behaviour (Lewinsohn et al. 1970; Libert & Lewinsohn, 1973; Teasdale et al. 1980). Studies of healthy mother–infant dyads have demonstrated that infants are highly sensitive to their interpersonal environment (Murray, 1988). It is, therefore, likely that postpartum depression, by virtue of its impact on maternal interpersonal functioning, will disrupt normal infant engagements with the mother and, as a consequence, impair infant developmental progress. In recent years a considerable body of research surrounding this question has accumulated.

Background. This paper describes the rationale and methodology of the first national psychiatric morbidity surveys to be carried out in Great Britain. The objectives of the surveys were to estimate the prevalence of psychiatric morbidity among adults aged 16–64 living in Great Britain; to identify the nature and extent of social disabilities associated with psychiatric morbidity; to describe the use of health and social services by people with psychiatric morbidity and to investigate the association between mental illness and potential environmental risk factors in a household sample.

Methods. Four separate surveys were carried out in order to meet the objectives; a private household sample (N=10108), a sample of institutions caring for the mentally ill (N=1191), a sample of homeless people (N=1166), and a supplementary sample of patients with psychosis living in private households (N=350). A two-stage assessment procedure was used, in which all subjects were given the Revised Clinical Interview Schedule (CIS-R) administered by lay interviewers to assess neurotic symptoms and disorders and a psychosis screen, including the Psychosis Screening Questionnaire. Those who were positive on the psychosis screen were then interviewed by psychiatrists using the SCAN (incorporating the tenth edition of the Present State Examination).

Conclusions. Large scale national surveys such as this augment the inadequate data on psychiatric morbidity that are routinely available and are, therefore, an important source of information upon which to base policy and generate aetiological hypotheses. These surveys provide a possible model for similar surveys in other countries.

Conclusions about medical interventions or the causes of disease are based upon reviews of the scientific literature. Single studies usually have limited statistical power or may be difficult to interpret or generalize and so the findings from a single study can rarely justify a change in clinical practice or in an aetiological theory. Even when planning larger studies or mega-trials (Yusuf et al. 1984), a thorough review of existing literature is needed and the results of the study need to be placed in that context, though single studies can exert an important and powerful influence.

Chronic fatigue syndrome (CFS) is characterized by severe and prolonged fatigue, affecting both physical and mental functioning, exacerbated by relatively minor exertion (Fukuda et al. 1994). A variety of other symptoms such as impaired concentration and memory, disturbed sleep, depressed mood and anxiety are also often present. Alongside this emerging consensus describing the clinical features of chronic fatigue, controversy has raged as to its aetiology, particularly the relative importance of viruses and other infectious agents, the contribution of neuromuscular abnormalities and whether the association with psychiatric disorders is primary or secondary.

Background. The authors examined data from a follow-up study of first admission schizophrenic patients treated with and without antipsychotic medications, who were discharged from the hospital within 6 months. It was predicted that patients who did not require antipsychotic medications for discharge would have a more favourable long-term outcome.

Methods. The subjects were part of the Camarillo State Hospital study conducted by May and colleagues in the late 1950s and early 1960s. Patients had been randomly assigned to treatment with and without antipsychotic medications. The number of rehospitalization days and total prescribed chlorpromazine equivalents were calculated for each patient for the 2 years following discharge. In order to assess patients' continuing ability to function, 11 patients from each group who met DSM-IV criteria for schizophrenia were matched for age, educational status at first admission, race, and gender; their Global Assessment of Functioning Scale (GAF) score was estimated across a period of 6–7 years following discharge.

Results. During the second year following discharge, patients initially treated with antipsychotic medications required fewer rehospitalization days than the initially non-medicated patients. Furthermore, 6–7 years following initial discharge, those patients initially treated with medications were functioning at a higher level, as measured by GAF scores, than patients not initially treated with antipsychotic medications.

Conclusions. The results of this study suggest that, at least for this subgroup of patients, early treatment with antipsychotic medications both decreases the immediate morbidity associated with schizophrenia, and prevents detrimental changes possibly related to prolonged untreated psychosis.

Background. Single photon emission tomography (SPET) with the dopamine D2/3 ligand 123I-IBZM gives a semi-quantitative estimate of dopamine binding. In depressed patients, we predicted evidence of reduced function, i.e. increased binding, particularly in more retarded patients.

Methods. Fifteen depressed patients with major depressive illness and 15 healthy, age- and sex-matched volunteers were examined with a clinical and neuropsychological test battery and high resolution IBZM-SPET. Estimates for specific binding were computed by averaging striatum to whole slice or frontal uptake ratios over 8–10 scans acquired from 70 min after tracer injection.

Results. Using whole slice as reference, left striatal uptake ratios did not significantly differ for patients from controls. Right ratios were significantly higher in patients than controls (P=0·03). There were significant correlations between IBZM binding in left and right striatum and measures of reaction time and verbal fluency.

Conclusions. Increased IBZM binding in striatum probably reflects reduced dopamine function, whether due to reduced release of dopamine, or secondary up-regulation of receptors. The observed abnormalities may be trait or state related, an issue that needs to be addressed with longitudinal study designs. The possible role of medication as a confounding variable requires further exploration.

Although claims that the eating disorders have reached epidemic proportions in the West are clearly exaggerated (Fairburn & Beglin, 1990), anorexia and bulimia nervosa are among the commonest psychiatric disorders to affect young women in Western societies (Whitaker et al. 1990; Kendler et al. 1991). Because of the high morbidity and mortality associated with these disorders (for reviews see Hsu, 1990, 1991), there is no cause for complacency, and continued research into their pathogenesis, treatment and outcome is needed.

Background. Patients with anorexia nervosa (AN) run a high risk of becoming chronically ill and of dying. In the acute phase of their illness they present with numerous physical and laboratory abnormalities. However, little is known about the long-term prognostic value of these findings or about the medical morbidity in large samples of consecutively treated patients in the long-term.

Methods. We evaluated 84 consecutive female patients with AN who were studied again an average of 11·9 years later. The ability of some of the laboratory data obtained at the initial examination to predict a fatal or chronic course was analysed by discriminant and multiple regression analyses.

Results. Abnormally low serum albumin levels ([les ] 36 g/l) and a low weight ([les ] 60% of average body weight) at the initial examination were variables best able to predict a lethal course. In addition, high serum creatinine and uric acid levels predicted a chronic course. Most of the initial abnormal laboratory findings were reversible with a normal food intake. At a 12-year follow-up 67% of the chronic anorexic patients and 27% of those in the good/intermediate outcome group (compared to 8% morbidity in normal controls) presented with medical co-morbidity, in particular osteoporosis and renal disease. The standardized mortality ratio was 9·6.

Conclusions. Laboratory findings obtained at the initial examination may be helpful in predicting a fatal or chronic course of AN. An evaluation of the long-term outcome of eating disorders should include an assessment of the medical co-morbidity.

Background. The nature and correlates of hypochondriasis are still poorly understood, especially in settings other than those in North America and Western Europe.

Methods. A total of 25916 consecutive patients making ambulatory visits to primary care clinics at 15 sites in 14 countries were screened using the 12-item General Health Questionnaire (GHQ-12). Based on the screen-score performance, a stratified sample of 5447 of respondents was evaluated at the second-stage (response rate: 62%). Evaluation consisted of physician-rated physical and psychological health status and self-ratings of overall health, physical disability and GHQ-28. Interviewers conducted assessment of psychiatric status, using the Composite International Diagnostic Interview (CIDI) and occupational disability.

Results. Across the sites, the occurrence of ICD-10 hypochondriasis was 0·8% (95% Confidence Interval, 0·5–1·0%) and over 1·5% at only two sites. A less restrictively-defined form of the disorder had a pooled frequency of 2·2% (95% Confidence Interval, 1·8–2·6%) across the sites. Patients with this abridged hypochondriasis were more likely than those without to have co-morbid major depression and generalized anxiety disorder. They had a poorer perception of their health, were more physically disabled, were more impaired in the performance of occupational role, and were above-average utilizers of health service. Patients with ICD-10 hypochondriasis were no more impaired than those with abridged hypochondriasis.

Conclusion. Even though the ICD-10-defined hypochondriasis is rare, a form consisting of the triad of disease conviction, associated distress and medical help-seeking is present in primary-care settings in different cultures. This syndrome is associated with considerable psychiatric ill-health and functional disability.

Background. Life events may play a role in precipitating the onset of bulimia nervosa. This study aimed to determine the rates of occurrence of certain life events during the year before onset of disordered eating in women with bulimia nervosa, and to compare them with the rates among age-matched normal controls.

Methods. The subjects were 102 women with bulimia nervosa and 204 women without an eating disorder, all recruited from the same community sample of young adult women. Interview measures were used for diagnosis and for assessment of life events.

Results. The bulimia nervosa cases reported more life events during the year before onset of disordered eating than were reported by controls of the same age. Events involving disruption of family or social relationships, or a threat to physical safety, were especially common among the bulimia nervosa cases.

Conclusions. These findings suggest that certain life events may play a role in precipitating the onset of bulimia nervosa. The study was retrospective in design. Prospective studies of the role of life events would be valuable, but difficult to perform.

Background. Gender differences in clinically relevant depression are well established, appear to be greatest in childbearing years and may be the result of gender differences in social roles.

Methods. A community sample of 100 couples who had recently experienced at least one threatening life event that was potentially depressogenic for both of them was studied using a semi-structured interviewer-rated interview. Onset of depression was assessed using the Present State Examination, and, rather than assuming that a gender difference in roles existed uniformly across the couples, they were characterized according to their actual role activity and commitment.

Results. Women were found to have a greater risk of a depressive episode following the life event than men, and this difference was of a similar magnitude to other reports of gender differences in depression. Consistent with a role hypothesis, this greater risk was entirely restricted to episodes that followed events involving children, housing or reproductive problems. In addition, it was found that women's greater risk of a depressive episode following such events was only present among those couples where there were clear gender differences in associated roles. There was some suggestion that differences in roles on the one hand resulted in women being more likely to hold themselves responsible for such events and, on the other hand, enabled men to distance themselves from them.

Conclusions. These results support the hypothesis that gender differences in rates of depression in the general population are, to a considerable extent, a consequence of role differences.

Background. This paper describes the Household Survey from the National Survey of Psychiatric Morbidity. This covered a sample drawn at random from the population of Britain, with the exception of the Highlands and Islands of Scotland.

Methods. The Postcode Address file was used as the sampling frame. Nearly 13000 adults aged 16–65 were selected for interview, of whom 10108 (79·4%) were successfully interviewed. Eight per cent could not be contacted and 13% refused interview. Psychiatric assessment was carried out by lay interviewers using the CIS-R. Subjects were also screened for psychosis, and screen-positive individuals were examined by psychiatrists using SCAN.

Results. Sixteen per cent of subjects scored above the standard cut-off of 12 on the CIS-R. The overall 1-week prevalence of neurotic disorder was 12·3% in males and 19·5% in females. Unmarried and post-marital groups had high rates of disorder, as did single parents and people living on their own. Respondents in Social Class I had notably lower rates of neurotic disorder than the remainder of the sample. Unemployment was strongly associated with disorder. Subjects living in urban areas had a higher overall prevalence, but there was no significant variation by region. Black respondents had higher rates of disorders that were entirely explained by their age, family type and social class. Individual neurotic disorders were all significantly commoner in women, with the exception of panic disorder. The 1-year prevalence of functional psychoses was 4 per 1000, with no sex difference. Alcohol and drug dependence was considerably more prevalent in men.

Conclusions. For the first time, the survey provides data on the prevalence and correlates of psychiatric disorder on a nationwide sample that can be used to inform equitable and effective national psychiatric services.

Background. Many studies have separately reported abnormalities of frontal and temporal lobe structures in schizophrenia, but little is known of structural fronto-temporal associations in this condition. We investigated whether male patients with chronic schizophrenia would show abnormal patterns of correlation between regional brain volumes.

Methods. Structural magnetic resonance images of the brain in 42 patients were compared with 43 matched unaffected controls. We explored the pattern of association between regional brain volumes by correlational analyses, and non-parametrically tested for significance of between-group differences by randomization.

Results. The schizophrenics demonstrated significant volume deficits in several brain regions (left temporal lobe and hippocampus, right dorsolateral prefrontal cortex), and significant volume increases in the ventricular system (third ventricle and left temporal horn of the lateral ventricle). Controls demonstrated large positive correlations (r>0·4) between prefrontal and temporal lobe regions. By contrast, inter-regional correlations significantly reduced in schizophrenics included those between prefrontal, anterior cingulate and temporal regions, and between posterior cingulate and hippocampus (P<0·05). The most salient abnormality in patients was a dissociation between prefrontal and superior temporal gyrus volumes (P<0·01).

Conclusions. These results support the existence of a relative ‘fronto-temporal dissociation’ in schizophrenia which we suggest may be due to lack of mutually trophic influences during frontal and temporal lobe development.

Background. The aim of this study was to determine differences between patients attending primary-care clinics with somatic symptoms for which no organic cause can be found, and those with a clear organic basis for their complaints.

Methods. Physical and psychiatric morbidity was measured in 5447 subjects presenting for primary care in 14 countries. Subjects were recruited using a two-stage design from 26969 consecutive attenders and 25916 (96·5%) subjects were successfully screened using the General Health Questionnaire (GHQ-12). The response rate for the second-stage assessment was 63% and consisted of the Composite International Diagnostic Instrument adapted for use in primary care (CIDI-PHC), the 28-item GHQ, the Groningen Social Disability Schedule (GSDS) and GP ratings to assess physical psychiatric and social status. Patients who reached the threshold for the number of somatic symptoms (five or more symptoms) were then divided into two groups on the basis of whether their somatic symptoms were medically explained or not.

Results. The presence of somatic symptoms, irrespective of aetiology, was associated with increased social and psychiatric morbidity. In the case of non-medically explained symptoms there was a clear linear relationship while with medically explained symptoms psychiatric morbidity rose sharply in the presence of more than 11 symptoms. Patients who had five or more non-medically explained symptoms were significantly younger, had greater psychiatric morbidity, were at greater risk of harmful use of alcohol and reported greater social disability than those with a medical explanation for their somatic symptoms. Patients who had an excess of somatic symptoms in both categories had especially high rates of social and psychiatric morbidity.

Conclusions. These results show a strong association between somatic symptoms irrespective of aetiology and psychiatric morbidity across disparate cultures.

Background. Previous work suggests that temporal lobe magnetic resonance imaging (MRI) can distinguish those with dementia of the Alzheimer type (DAT) from healthy age-matched controls. However, its specificity with regard to conditions such as vascular dementia, depression and other disorders associated with cognitive impairment has not been determined.

Methods. We studied 222 subjects using T1 weighted MRI with 5·1 mm coronal slices throughout the temporal lobe. Subjects included: healthy controls (N=40); DSM-III-R major depression (N=61); NINCDS/ADRDA DAT (N=77) and OTHER (N=44, comprising subjects with vascular dementia, Huntington's disease, schizophrenia, alcohol related cognitive impairment and a group of ‘memory complainers’). Hippocampus, amygdala, entorhinal cortex, parahippocampal gyrus and cerebral cortex were rated visually on a 0–3 scale by two experienced neuroradiologists blind to clinical diagnosis.

Results. Ratings of temporal lobe atrophy provided good separation between those with AD and all other groups. For example, anterior hippocampal atrophy had a sensitivity of 83% for detecting DAT, a specificity of 80% for controls, 87% for depressed subjects and 89% for OTHER. Other regions were less sensitive, but more specific for the diagnosis of DAT. In particular parahippocampal gyrus and entorhinal cortex had high specificity (97% for depressed subjects and 98% for OTHER). Because of an age-related increase in atrophy, sensitivity was highest for those over the age of 75, while specificity was highest for younger subjects. Significant correlations were observed between atrophy ratings of hippocampus, amygdala, entorhinal cortex and parahippocampal gyrus and CAMCOG memory score and length of history.

Conclusions. Temporal lobe MRI may have an important role in assisting with the clinical diagnosis of DAT, particularly its differentiation from depression and other disorders that may cause diagnostic difficulties in clinical practice.