We examined the differences in volume of the ventricular and extracerebral cerebrospinal fluid spaces in normal ageing and in probable Alzheimer's disease (AD) and we tried to investigate the effects of the severity of illness on the morphometric differentiation of AD and ageing, the principal components underlying brain atrophy in both conditions, and the correlations of these measurements with clinical findings.

Forty patients with probable AD were matched with 40 non-demented elderly controls. Both groups underwent standardised clinical tests and unenhanced cranial computed tomography for post hoc volumetric analysis.

The lateral and third ventricles and the anterior and lateral fissures were significantly larger in AD than in normal ageing. The volumes of the lateral ventricle and lateral fissure permitted a highly efficient differentiation between normal ageing and AD even at the mild stage of dementia, and this differentiation was improved further in the more severe stages of illness. We identified one principal component underlying brain atrophy in normal ageing and two components in AD: a ‘grey matter’ component accounting for sulcal and third ventricular enlargement, and a ‘white matter’ component for lateral ventricular enlargement. In AD, most of the volumetric measurements were significantly correlated with cognitive impairment, but in the group of non-demented elderly controls they were correlated with age.

Volumetric indices of brain atrophy permit a highly efficient differentiation between normal ageing and AD even in the mild stages of illness and this demonstrates that substantial structural brain changes have developed in the preclinical phase of illness. We suggest that there is an uncoupling between lateral ventricular enlargement and cortical brain atrophy in AD.

Little is known about the impact of personality pathology on the treatment outcome of major depressive illness in primary care in the UK.

Patients meeting criteria for DSM–III major depressive disorder were randomly allocated to one of four treatments each lasting 16 weeks, then followed up for 18 months. Assessments were made of depressive symptoms, personality and social functioning. Personality was assessed at maximum improvement or 16 weeks.

The prevalence of personality disorder (PD) in the sample of 113 patients was 26%. Patients with a PD were significantly younger and rated more depressed at entry than patients with no personality disorder (NoPD). On completion of treatment patients with a PD were significantly more depressed and had poorer social functioning than the NoPD group. After 18 months there were no differences in ratings of depression or social functioning between the groups.

There was substantial improvement in both the PD and NoPD groups. The presence of personality pathology delays recovery from major depressive illness.

It was hypothesised that schizophrenic patients with tardive dyskinesia show an excess of neurodevelopmental disturbance, particularly minor physical anomalies, in association with cognitive dysfunction and abnormalities of cerebral structure.

Forty-seven out-patients with a DSM–III diagnosis of schizophrenia were examined for tardive dyskinesia using the Abnormal Involuntary Movement Scale; they were examined also for minor physical anomalies and neuropsychological test performance. Cortical atrophy, signal hyperintensities and lateral ventricular volume were determined on magnetic resonance imaging.

Patients with and without tardive dyskinesia could not be distinguished by age, gender distribution or a number of clinical measures; however, patients with tardive dyskinesia sorted fewer categories on the Wisconsin Card Sorting Test (P = 0.04). Cerebral structure in patients with and without tardive dyskinesia could not be distinguished on magnetic resonance imaging but those with dyskinesia, all of whom showed involvement of the orofacial region, showed more evident minor physical anomalies of the head relative to those of the periphery (P = 0.02).

Tardive orofacial dyskinesia in schizophrenia appears to be associated particularly with poorer frontal lobe function, while predominance of craniofacial dysmorphogenesis may constitute a vulnerability factor that is related to the early origins of the disease process.

The study objective was to investigate the age-related variation of social factors in suicide.

Age-related variations in marital status, living arrangements, employment and social interaction were investigated in an entire 12-month suicide population in Finland (n = 1067); these findings were compared with appropriate census data.

Several social factors varied across age groups among suicides, with some age-related sex differences. Compared with the general population, the suicides were more commonly never married (especially men aged 30–39 years), divorced, and widowed (especially women aged 60–69 years); living alone was more frequent among the suicides, as was living with parents among male suicides aged 25–39 years. A history of psychiatric admission was especially common among young male suicides who had never married or were living with parents. Living alone was particularly frequent among middle-aged male suicides who had misused alcohol.

While most of the age-related variation in social factors found in suicide seems to parallel the natural variation of these factors in the general population, some social findings in suicide might be related to the victims' psychopathology and excessive alcohol use.

This paper describes the 13-year outcome of an epidemiologically defined and representative cohort of patients selected when they were experiencing their first episode of schizophrenia.

In a 13-year follow-up study of a cohort identified in Nottingham in 1978–80, the outcome (symptoms, disability, residence and treatment) was assessed using standardised instruments.

Four of the original 67 patients with ICD–9 schizophrenia were lost to follow-up and five were dead: 52% were without psychotic symptoms in the last two years of follow-up, 52% were without negative symptoms and 55% showed good/fair social functioning. However, only 17% were alive at follow-up, without symptoms and disability, and receiving no treatment.

The findings reported are similar to those of other long-term follow-up studies of schizophrenia and also to 5-year follow-up studies. Kraepelin's emphasis on the longitudinal implications of a diagnosis of schizophrenia are supported, but may be over-pessimistic.

In the search for population-based indicators of need for mental health services, psychiatric admission rates have been correlated with sociodemographic variables. We explored such correlations for different diagnostic groups.

Admissions data for the 19 districts in the North West Region were derived from the Korner Episode System for 1992/3 and divided into eight broad diagnostic groups using ICD–9 codes. Admission rates per 1000 were correlated with measures of deprivation derived from the 1991 census data and with standardised mortality ratios. For the two largest diagnostic groups, correlations with age-standardised admission rates were also calculated.

For schizophrenia/delusional disorder, eight of the 10 sociodemographic measures were significantly correlated with admission rates (Pearson's r 0.52–0.79). On all measures these correlations were greater than those seen for total mental illness. Significant positive correlations of a lower order were seen for organic brain syndromes and mania. Admission rates for depression, personality disorder and substance misuse were not significantly correlated with any of the sociodemographic measures. Admission rates for neurotic illness were negatively correlated with all deprivation measures, with the negative correlation statistically significant at the 5% level for ethnic composition and overcrowding. Standardising admission rates for age and repeating the analysis after removal of influential data points did not greatly alter these findings.

The association between psychiatric admission rates and measures of deprivation varies considerably with diagnosis. Measures of social deprivation may indicate need for services for patients with psychotic disorders; admission rates for non-psychotic illnesses may reflect the availability of beds rather than need.

Paracetamol self-poisoning, which carries a significant risk of fatal liver damage, is increasing in the UK, especially among adolescents. There is concern that media portrayal of suicidal behaviour may influence its occurrence. We have investigated the effects of two broadcasts of a television drama showing a teenage girl's overdose of paracetamol.

Data from the Oxford Monitoring System for Attempted Suicide were used to examine changes in the number of overdoses and the use of paracetamol in the three week period before and after both broadcasts. An analysis of log-linear models was carried out, using additional data from two previous years, in order to control for the effects of gender, age group, time period, season, year and drug. A questionnaire was used to ascertain whether the programme had influenced patients' decisions to take an overdose or their choice of drug.

An increase in paracetamol overdoses occurred following the first broadcast, but when data from the two previous years were examined and the log-linear analysis was used there was no evidence of significant effects attributable to the broadcasts. The questionnaire study revealed that very few people had seen either episode, and only two claimed that it had influenced them (one in terms of getting help).

It is important to use sound methodology to avoid interpreting chance fluctuations as a media effect. The potential positive benefits of such programmes should also be investigated.

A previous cross-national epidemiological study of first admission rates in London and in Aarhus, Denmark, found that the incidence of mania was virtually identical for both centres. This study sought to examine the corresponding rate for a defined catchment area in Dublin, Ireland, and to establish whether the impression of a higher rate could be validated.

The study combined a six-year retrospective review and one-year prospective collection of first-admission cases of mania from a defined catchment area. Diagnosis was according to ICD criteria and the Syndrome Check List of the Present State Examination (PSE). Cross-national comparisons were made using standardised incidence rate ratios (SIRs).

The crude incidence rate for Dublin in the age range 18–60 years was established at 4.5 per 100 000 per year. A comparison of SIRs for the three centres showed the Dublin rate to be higher than expected (P < 0.02). Age-specific analysis indicated that this increase came mainly from the age range 18–29 (χ2 = 9.08, P = 0.01).

The study confirmed the impression that the local incidence of mania in Dublin is higher than that reported from two other North European centres. A variation in rates from contrasting socioeconomic districts within the catchment area points to some caution in the interpretation of the results and suggests further study from wider catchment areas.

Gender may have a significant impact on the prevalence, age at onset, symptoms, course and outcome of schizophrenia, as well as on the pattern of psychopathology in relatives.

We examined these questions in the Roscommon Family Study, in which the probands were epidemiologically sampled from a case registry and followed up an average of 15 years after onset. Face-to-face interviews were conducted with 86% of traceable living relatives.

The treated lifetime prevalence of DSM–III–R schizophrenia was 0.54 ± 0.06% in men and 0.28 ± 0.04% in women. No significant differences were seen in the age at onset, symptoms, course or outcome of schizophrenia. The risks for schizophrenia, schizophrenia spectrum disorders, affective illness and alcoholism were similar in relatives of male and female schizophrenic probands.

Gender has little impact on the presentation and course of schizophrenia in the west of Ireland. The familial liability to schizophrenia did not differ in affected men and women. No evidence was found that schizophrenia in women, compared to men, is, from a symptomatic or familial perspective, more closely related to affective illness. The substantial gender difference in the prevalence rate of schizophrenia in Ireland cannot be explained by women having a greater resistance to the familial predisposition to illness.

Although poor prognosis has been considered a defining characteristic of schizophrenia, long-term studies show marked heterogeneity of outcome.

Assessments of positive and negative symptoms, premorbid and current IQ, and months of in-patient care made in an outcome study of 342 schizophrenic patients were categorised by severity. Determinants of these categorisations were sought from the historical variables available, using analysis of variance. Vignettes of patients with the best and worst symptomatic outcomes were then compared.

Negative symptoms were associated with early onset, male sex and poor academic record. Positive symptoms were associated with occupational decline. Cognitive decline was associated with occupational variables, and in-patient care with academic and occupational variables. The vignettes showed that good outcome was associated with family psychiatric history and poor outcome with unavailability of family history.

The findings support the view that the most malignant form of schizophrenia is neurodevelopmental, but poor outcome was clearly associated with family fragmentation.

Many patients complain less of their auditory hallucinations per se than of lack of control of the experiences. There is reason to believe that a non-distraction (exposure) approach could help patients gain more control over persistent auditory hallucinations and teach them that their experience is a form of thinking and has no external source. This study is a pilot test of that idea.

Five DSM–III–R schizophrenic outpatients with medication-resistant auditory hallucinations improved with a mean of 31 hour-long sessions over 3 months of therapist-guided exposure to their hallucinations and situations likely to evoke them.

Improvement was greatest in patients' anxiety and sense of control over their hallucinations, less in social use of leisure and hallucinating time.

These mildly encouraging pilot results warrant a controlled study of exposure for drug-resistant chronic auditory hallucinations and other psychotic experiences which are associated with anxious avoidance.

This study compared long-term outcome of serious mental illness in two states using a matched design to assess psychiatric rehabilitation programmes; Vermont subjects participated in a model psychiatric rehabilitation programme, while the Maine group received more traditional care.

Maine and Vermont subjects (n = 269) were matched by age, sex, diagnosis, and chronicity. Demographic, illness, and life history information were abstracted from hospital records by clinicians blind to outcome. DSM–III criteria were applied retrospectively. Outcome was assessed by clinicians blind to history.

Vermont subjects alive at follow-up (n = 180) were more productive (P < 0.0009), had fewer symptoms (P < 0.002), better community adjustment (P < 0.001) and global functioning (P < 0.0001) than Maine subjects (n = 119).

Outcome differences may be due to Vermont's model programme and a policy of allowing an earlier opportunity for community life.

While neuropsychological deficits are recognised to occur in manic-depressive psychosis during episodes of depression and to reverse with clinical recovery, it is uncertain whether they can ever be seen outside episodes of illness.

Forty-five patients meeting DSM–III–R criteria for major depression or bipolar disorder were screened using tests of memory, executive function and overall intellectual function. All testing was carried out during remission of affective symptoms.

None of 24 young patients and 11 elderly patients scored in the impaired range on any of the tests. However, five of ten patients with chronic, severe affective disorder were impaired on one or more of the measures. On more detailed neuropsychological investigation, these five patients were found to show a variable pattern of impairment, ranging from memory and executive deficits in relative isolation, to widespread poor performance.

Enduring neuropsychological deficits may be a feature of chronic, severe manic-depressive illness.

Afro-Caribbean immigrants are reported to have a high rate of schizophrenia compared with other population groups.

In a prospective first contact study of schizophrenia in Jamaica in 1992, 335 patients were examined using the Present State Examination.

285 patients were evaluated as having a PSE ‘restrictive’ S + diagnosis of schizophrenia, and 32 as having a ‘broad’ S?, P, or O diagnosis of schizophrenia. With a population of 2.46 million, this represents a first-contact incidence rate for ‘restrictive’ schizophrenia of 1.16 per 10 000 population, and an age-corrected (15–54) incidence rate of 2.09 per 10 000.

Incidence rates for schizophrenia in Jamaica are lower than those reported in Afro-Caribbean immigrants in the UK and Holland, and within the reported range for other population groups worldwide.

Clozapine's effectiveness in reducing symptoms and facilitating discharge among patients with chronic schizophrenia who were resistant to neuroleptics was studied.

All 169 such patients in a public psychiatric hospital were given clozapine. BPRS ratings (0–5 scale) were completed before treatment and 21 months later. Patients were followed for about 2.5 years.

Clozapine was discontinued in 37.8% of cases due to non-compliance, non-response, or side-effects. At follow-up 41 % of clozapine recipients and 25.9% of the drop-outs were discharged and remained so, and 33% of recipients and 24.1% of drop-outs were being prepared for discharge. Longer treatment was associated with more improvement. Decline in average BPRS total scores of recipients was significantly more than drop-outs (32.7, s.d. 16.8 v. 12.1, s.d. 14.1, d.f. = 155, t = 7.5, P = 0.000).

Clozapine appears to be effective for treating some chronic neuroleptic non-responding schizophrenic patients.