Risk factors of depression in later life, particularly for sub-cases and for psychotic and neurotic types of depression, are unclear.

To identify such risk factors.

Over 5200 older people ($65 years), randomly selected from Liverpool, were interviewed using the Geriatric Mental State (GMS)and the Minimum Data Set (MDS). The computer-assisted diagnosis AGECAT identified 483 cases and 575 sub-cases of depression and 2451 with no mental problems. Logistic regression was employed to examine factors relevant to caseness.

In multiple logistical regression, odds ratios (ORs) were significantly high for being female (2.04, 95% CI 1.56–2.69), widowed (2.00, 1.18–3.39), having alcohol problems (4.37, 1.40–2.94), physical disablement (2.03, 1.40–2.94), physical illness (1.98,.1.25–3.15), taking medications to calm down (10.04, 6.41 −15.71), and dissatisfaction with life (moderate 4.54, 3.50–5.90; more severe 29.00, 16.00–52.59). Good social networks reduced the ORs. If sub-cases were included as controls, the statistical significance was reduced.

Age was not associated with depression in later life whereas gender, physical disablement and dissatisfaction with life were. The sub-cases shared many risk factors with cases, suggesting that prevention may need to be attempted at an early stage.

Social cognitive skills are those which enable understanding of social situations; they are relevant to a variety of psychiatric disorders including autism, schizophrenia and externalising behaviour problems in children.

To examine the heritability of social cognitive skills.

Using a population-based sample of twins aged 5–17, the genetic and environmental influences on social cognitive skills were examined.

Male scores were higher than female scores (P < 0.001), indicating poorer social cognition among males. A heritability of 0.68 (95% CI 0.43–0.78) was found, with shared environmental influences accounting for only 0.05 of the variance (95% CI 0.00–0.28). This could be removed from the model without worsening the fit. There were no significant differences in genetic effects between the genders, but age-related changes were found, with younger twins showing greater genetic influence on social cognition.

Social cognition appears to be under considerable genetic influence in the population and shows significant male–female differences. No gender differences in genetic influences on the variance of scores were found, but the effects of age were significant.

The meaning of post-stroke depression is controversial.

To investigate the hypothesis that major post-stroke depression (PSD) may be due to organic factors (left frontal lesions) immediately after the stroke, but to psychosocial factors in later stages.

We studied 153 consecutive stroke patients, categorised on the basis of time elapsed since stroke, lesion location and presence/absence of major PSD, Fifty-eight were examined in the first two months following the stroke, 52 between two and four months, and 43 after four months or more. The symptom profiles and anatomical–clinical correlates of major PSD were studied in each subgroup. A group of 30 patients affected by a functional form of major depression were also investigated.

The symptom profiles and anatomical–clinical correlates of major PSD were not different in the acute and more chronic stages. Clear symptom differences were, however, observed between major PSD and endogenous major depression. Motivated (reactive) symptoms prevailed in the former, whereas unmotivated symptoms prevailed in the latter.

Our data are more consistent with a psychological than with a neurological model of post-stroke depression.

The Hospital Anxiety and Depression Scale (HADS), developed for use with adults, has potential with adolescents.

To test the HADS's validity with adolescents.

The HADS was given to 248 schoolchildren and to 48 psychiatric outpatients and 38 deliberate self-harm inpatients aged 12–17, and validated against ICD–10 diagnoses for the out-patients.

The HADS has adequate test-retest reliability and factor structure, and discriminates between adolescents diagnosed with depressive or anxiety disorders and those without these diagnoses.

The psychometric properties of the HADS and its shortness make it useful for screening and in clinical settings with adolescents.

The use of MDMA (‘ecstasy’) is common among young people in Western countries. Animal models of MDMA toxicity suggest a loss of serotonergic neurons, and potentially implicate it in the development of significant psychiatric morbidity in humans.

To test whether long-term use of MDMA can produce abnormalities in cerebral serotonin, but not dopamine, transporter binding measured by single photon emission computed tomography (SPECT)

Ten male regular ecstasy users and 10 well-matched controls recruited from the same community sources participated in SPECT with the serotonin transporter (SEPT) ligand [123I]-CIT. Dopamine transporter binding was determined from scans acquired 23 hours after injection of the tracer.

Ecstasy users showed a cortical reduction of SERT binding, particularly prominent in primary sensory-motor cortex, with normal dopamine transporter binding in lenticular nuclei.

This cross-sectional association study provides suggestive evidence for specific, at least temporary, serotonergic neurotoxicity of MDMA in humans.

It has been suggested that low serum total cholesterol is associated with an increased risk of suicide.

To study the association between serum total cholesterol, depression and suicide using versatile, prospective data.

A total of 29 133 men aged 50–69 years were followed up for 5–8 years. Baseline blood samples were analysed for serum total and high-density lipoprotein cholesterol concentrations. Self-reported depression was recorded, data on hospital treatments due to depressive disorders were derived from the National Hospital Discharge Register and deaths from suicide were identified from death certificates.

Low serum total cholesterol was associated with low mood and subsequently a heightened risk of hospital treatment due to major depressive disorder and of death from suicide.

Our results suggest that low serum total cholesterol appears to be associated with low mood and thus to predict its serious consequences.

This study evaluates two in-patient units attached to community mental health centres (CMHCs) that were designed to supplement acute in-patient care and to integrate with community-based after-care services.

To examine the comparative outcome of patients with severe mental illness (SMI) admitted to the two units.

All patients with SMI admitted to the acute psychiatric wards serving the two CMHCs, those transferred to the community in-patient units and those admitted directly to these units (n=110) were compared with patients (n=67) admitted to acute wards serving two similar catchment areas without associated community beds. Baseline clinical and social measures were made and repeated at six and 12 months. Satisfaction with services was assessed at 12 months.

The experimental group showed significantly better outcomes, significant reduction in unmet need and better satisfaction with services.

The use of the community beds appears to have significant benefits for patients with SMI.

In the past two decades the suicide rate in young people has risen in the UK and other countries.

To identify the characteristics of people aged under 35 who commit suicide.

Psychological autopsy study with case–control design. Cases were 84 suicides and probable suicides (open verdicts). Controls were 64 age- and gender-matched non-suicides obtained through the general practices of the matched cases. Main variables: social, including employment status, marital status and living circumstances; life events and interpersonal difficulties; clinical, including current psychiatric disorder, alcohol and drug misuse, previous self-harm and personality disorder.

A large number of highly significant social, interpersonal and clinical differences between suicides and controls was found. Factor analysis identified two groups of variables within each of the social, interpersonal and clinical domains which were independently linked to suicide, corresponding to: acute, severe mental disorder and chronic disorder of behaviour; rootlessness and social withdrawal; chronic and recent interpersonal problems.

A broad prevention strategy is needed to reverse the recent rise in the suicide rate in young people.

Family relationships in the transition from childhood to adult life may be important mediators of risk and resilience for adult psychopathology.

To develop a reliable and valid measure of the quality of relationships with each parent in young adults.

Interrater reliability of the Relationship with Family of Origin Scale (REFAMOS) was assessed from audio-taped interviews with 59 subjects. Age-related trends and associations with recalled childhood relationships were examined in survivors of childhood cancer and their controls (n=178).

Intraclass correlation coefficients were in the range 0.69–0.95, and κ values 0.80–0.82. Indices of current closeness to mothers were negatively correlated with age of subject and positively correlated with recalled maternal care in childhood. Negative qualities in current relationships were correlated with recalled overprotection.

The REFAMOS has good interrater reliability, and shows the predicted age-related trends in scores and associations with recalled childhood relationships.

This paper reports on the follow-up of a cohort of parentally bereaved adults with learning disabilities.

To investigate whether significant psychopathology, present up to 2.1 years after the death, had resolved five years later.

Of an original sample of 50 adults with learning disabilities, 41 were reassessed. The Aberrant Behaviour Checklist and the Psychopathology Instrument for Mentally Retarded Adults were re-administered to carers.

At follow-up, there was a small increase in the measures of aberrant behaviour. Measures of psychopathology showed improvement, and in particular there was a reduction in anxiety.

The response to bereavement by adults with learning disabilities is similar in type, though not in expression, to that of the general population. Learning disability is a significant predictor of mental health problems following bereavement. Participants adapted more easily when basic emotional needs had been constructively met by carers.

In a recent study, 20% of the variance in the weight of infants of mothers with eating disorders was accounted for by mealtime conflict.

To investigate the antecedents and interactive processes involved in the development of such conflict.

Mothers with eating disorders and their 12-month-old infants (n=34) and a comparison group (n=24) were videotaped during infant mealtimes. Specific classes of antecedents to conflict episodes were identified. An examination was then made of all such antecedents not leading to conflict and the mother–infant responses to each antecedent.

Within the index group, conflict was less likely when mothers acknowledged infants' cues and were able to put aside their own concerns. The relationship between maternal responses and the evolution of conflictual interaction was confirmed in multiple regression analyses including both index and comparison groups.

Mothers' and infants' responses to potential antecedents to conflict had an impact on whether mealtime conflict ensued. Conflict arose because maternal eating disorder psychopathology interfered with aspects of responsive parenting.

The influence of the macro-economic climate on suicide is unclear. During the recent recession, rates have increased in young males but declined in females.

To investigate associations between unemployment and suicide in 15 – to 44-year-old men and women over a period spanning two major economic recessions (1921–1995). To minimise confounding by changes in method availability, analyses are restricted to suicides using methods other than poisons and gases.

Time-series analysis using routine mortality and unemployment data.

There were significant associations between unemployment and suicide in both males and females. Associations were generally stronger at younger ages.

Secular trends in youth suicide may be influenced by unemployment or other factors associated with changes in the macroeconomic climate. These factors appear to affect women to the same extent as men. Although it is not possible to draw firm aetiological conclusions from time-trend data, our findings are in keeping with those of person-based studies.

Suicide rates in youngmales in the UK have risen markedly in recent years.

To investigate the characteristics of a series of consecutive suicides in under-25-year-olds.

We studied coroners' inquest notes, general practitioners' records and psychiatric case notes of 174 individuals (148 males and 26 females) whose deaths received a verdict of suicide or an open or accidental verdict (excluding traffic accidents) where the circumstances strongly suggested suicide.

More individuals were of lower social class and unemployed than in the local population. Hanging and carbon monoxide poisoning were the most frequent methods of suicide, and co-proxamol was the drug most often used in overdoses. Previous self-harm had occurred in 44.8%, nearly half of these having carried out multiple episodes and 80% having self-harmed within the previous year. Little support was found for an earlier finding of increasing frequency of general practitioner visits shortly before death. Only 22.4% of individuals were in the care of psychiatric services.

Diverse strategies are required to prevent suicide in the very young.

The vast majority of suicides suffer from at least one mental disorder at the time of death.

To identify risk factors for suicide, particularly those independent of current DSM–III–R Axis I disorder(s)

A case–control psychological autopsy study comparing suicides with matched community controls.

Independent risk factors for suicide included: Axis II (personality) disorder (particularly antisocial, avoidant and dependent); at least one of 12 life events (from the List of Threatening Experiences) during the previous 52 or 4 weeks (in particular, a ‘serious problem with close friend, neighbour or relative’); current unemployment; previous history of deliberate self-harm; and contact with a GP within 26 weeks. Relative to individuals with no current mental disorder, the estimated risk of suicide in those with Axis I–Axis II comorbidity (OR 346.0) was significantly greater than that in those with Axis I disorder(s) only (OR 52.4)

Suicide risk assessment may be enhanced by enquiry about the aforementioned independent risk factors, and attention to Axis I–Axis II comorbidity.

The clinical assessment and management of the risk of violence and suicide by people with mental illness may have to focus on environmental as well as individual factors.

To investigate possible associations between violence, homicide and suicide rates, population density and indices of deprivation, with particular reference to inner-city boroughs.

Coroners' statistics in London for homicide and suicide were obtained, with police-reported homicide and violence rates as a validity check. Correlations were made between these data and population density, the Jarman under-privileged areas score, and the Mental Illness Needs Index for each of the 32 London boroughs.

Homicide rates had a 14.3-fold range, suicide a 4.4-fold range and interpersonal violence a 6.6-fold range. The variables under study were strongly correlated with each other. Rates were highest in boroughs with high population density and deprivation scores. The associations persisted when covarying for deprivation, age structure or population density.

Because violence, homicide and suicide are so closely correlated, they are likely to be valid indices of the differences between adjacent boroughs; this has implications for the delivery of preventive and mental health services and for clinical management of risk.

Innovative approaches to the provision of psychiatric care must justify their ability to improve the quality of life within the resource constraints imposed on psychiatry.

To examine the average costs per patient of the experimental and control group services.

An individual patient costing methodology that identified, measured and valued all public and private resources.

The experimental group was more likely to remain in contact with services over a 12-month period, had fewer acute readmissions and spent less time in acute in-patient units. There were significantly different levels and patterns of resource consumption between the groups and between the two separate catchment areas.

The cost analysis should be assessed in the context of the previous outcome analysis. It is likely, but not inevitable, that such units will increase the overall costs of care provision; this largely depends on the effectiveness with which such units are integrated into existing care provision.

Previously reported results have demonstrated the efficacy of exposure and cognitive therapy in the treatment of chronic post-traumatic stress disorder (PTSD), but have not shown one to be superior to the other.

To investigate whether treatment benefits and equivalence are maintained at 12-month follow-up in patients with chronic PTSD treated with either imaginai exposure or cognitive therapy.

Twelve-month follow-up of a randomised clinical trial.

Fifty-four subjects (87% of the sample) were available to follow-up. They did not significantly differ clinically from drop-outs. There was significant clinical improvement at 12 months compared with pre-treatment. However, 39% of those followed-up still met criteria for PTSD. There were no significant differences between the two treatments. Victims of crime displayed higher levels of symptoms at follow-up than victims of accidents.

Clinical benefits for exposure or cognitive therapy were maintained.

Subsequent to initial exposure to the use of a psychoactive substance, psychoactive substance use disorder (PSUD) may or may not develop.

To investigate the relationship between the risk factors for initiation and the subsequent misuse of psychoactive substances.

The lifetime history of illicit substance use and misuse was obtained by telephone interview with 1934 members of female–female twin pairs. We apply a novel model, which estimates the role of genetic and environmental risk factors that influence initiation and those specific to misuse, to three classes of illicit psychoactive substances.

The individual-specific environment and family environment influenced the probability of initiation, but only individual-specific environment had an impact on the probability of subsequent misuse. Genetic factors which influence the risk of initiation and of misuse were identified.

Aetiological factors that influence drug initiation and subsequent misuse are correlated but not identical. Family environment is an important determinant of risk for drug experimentation. Two classes of genetic risk factors act on the liability to PSUD: those that influence the probability of initiation and those that influence the risk of misuse.

Axis I comorbidities are prevalent among patients with severe bipolar disorder but the clinical and psychopathological implications are not clear.

To investigate characteristics of four groups of patients categorised as follows: substance abuse only (group I), substance abuse associated with other Axis I disorders (group 2), non-substance-abuse Axis I comorbidity (group 3), no psychiatric comorbidity (group 4)

Consecutive patients with bipolar disorder with psychotic features (n=125) were assessed using the Structured Clinical Interview for DSM–III–R – patient version, and several psychopathological scales.

By comparison with group 4, group I had a higher risk of having mood-incongruent delusions, group 2 had an earlier age at onset of mood disorder, a more frequent onset with a mixed state and a higher risk of suicide, and group 3 had more severe anxiety and a better awareness of illness.

Substance abuse, non-substance-abuse Axis I comorbidity and their reciprocal association are associated with different characteristics of bipolar disorder.