The selective serotonin reuptake inhibitors (SSRIs) are now widely used in the treatment of depressive illness. Interest has grown in the use of SSRIs as alternatives to tricyclic antidepressants (TCAs) and in the therapeutic use of combinations of SSRIs and TCAs in refractory depression.

MEDLINE and PSYCLit literature searches were conducted. Reference sections from papers retrieved were scrutinised for other relevant reports.

Of 41 relevant articles identified, 35 were selected for review.

Fluoxetine, fluvoxamine, paroxetine and sertraline may substantially increase TCA plasma levels when given concurrently. Such interactions may give rise to adverse effects. The effect of sertraline may be less profound than that of fluoxetine, fluvoxamine and paroxetine. Limited data suggest that citalopram may not affect TCA serum levels. There is scant literature evidence to support the use of SSRIs in combination with TCAs as a treatment for refractory depression.

Psychosocial factors may contribute 25–30% to the outcome variance in bipolar disorders. Sufferers have identified benefits from psychotherapy, but biological models and treatments dominate the research agenda. The author reviews research on psychosocial issues and interventions in this disorder.

Research on adjustment to the disorder, interpersonal stressors and obstacles to treatment compliance were located by computerised searches and the author's knowledge of the literature. All published outcome studies of psychosocial interventions in bipolar disorder are reviewed.

There is an inadequate database on psychosocial factors associated with onset and maintenance of bipolar disorder. While the outcome studies available are methodologically inadequate, the accumulated evidence suggests that psychosocial interventions may have significant benefits for bipolar sufferers and their families.

Given the significant associated morbidity and mortality, there is a clear need for more systematic clinical management that addresses psychosocial as well as biological aspects of bipolar disorder. The author identifies appropriate research strategies to improve knowledge of effective psychosocial interventions.

People with severe mental illness often have a complex mixture of clinical and social needs. The Camberwell Assessment of Need (CAN) is a new instrument which has been designed to provide a comprehensive assessment of these needs. There are two versions of the instrument: the clinical version has been designed to be used by staff to plan patients' care; whereas the research version is primarily a mental health service evaluation tool. The CAN has been designed to assist local authorities to fulfil their statutory obligations under the National Health Service and Community Care Act 1990 to assess needs for community services.

A draft version of the instrument was designed by the authors. Modifications were made following comments from mental health experts and a patient survey. Patients (n = 49) and staff (n = 60) were then interviewed, using the amended version, to assess the inter-rater and test-retest reliability of the instrument.

The mean number of needs identified per patient ranged from 7.55 to 8.64. Correlations of the inter-rater and test-retest reliability of the total number of needs identified by staff were 0.99 and 0.78 respectively. The percentage of complete agreement on individual items ranged from 100–81.6% (inter-rater) and 100–58.1% (test-retest).

The study suggests that the CAN is a valid and reliable instrument for assessing the needs of people with severe mental illness. It is easily learnt by staff from a range of professional backgrounds, and a complete assessment took, on average, around 25 minutes.

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.

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.

Tardive dyskinesia (TD) may be mediated through free radical damage to neurons. Plasma lipid peroxide levels are a measure of radical damage to fats. Vitamin E is a free radical scavenger.

One hundred and twenty-eight schizophrenic patients were examined for TD using the Abnormal Involuntary Movements Scale. Blood samples were taken to measure plasma lipid peroxide, serum vitamin E and cholesterol, and vitamin Exholesterol ratios. Twenty-four patients were also examined in October 1993, January 1994, and April 1994. Biochemical results were compared in 81 patients and 79 normal subjects.

Patients with and without TD did not differ in median plasma lipid peroxide and serum vitamin E levels, or vitamin Exholesterol ratios. Correlations between seasonal change scores in TD and biochemical measurements were low. Lipid peroxide levels were higher and vitamin Exholesterol ratios lower in patients than in normal subjects. Vitamin E levels were lowest in in-patients and in those living in supported accommodation.

The results do not support the hypothesis that TD is mediated through free radical damage to neurons, but suggest increased free radical activity in schizophrenia.

This study examined the effect of emotional salience on the severity of thought disorder in schizophrenic patients.

Ten thought disordered and ten non-thought disordered schizophrenic patients were interviewed under two conditions: a personal interview involving material which was emotionally salient and an impersonal interview involving material which was not emotionally salient.

Both groups exhibited some thought disorder during both interviews. The thought disordered patients exhibited significantly more thought disorder during the emotionally salient interview.

Thought disorder in schizophrenic patients is affected by the emotional salience of the material being discussed. Clinical implications are discussed.

Insight has recently re-emerged as an important aspect of psychopathology amenable to empirical study. We sought to examine the relationship between various aspects of insight into illness and clinical, sociodemographic and neuropsychological variables.

From an inner-London catchment area population, 150 in-patients with recent onset of psychosis were assessed on a variety of measures, including the Present State Examination (PSE). Subjects were followed up for a mean of four years and reassessed.

High IQ was associated with better insight as rated on the PSE, while gender, ethnicity and a diagnosis of schizophrenia appeared to be unrelated. At follow-up, similar associations were found, as well as correlations with attitudes to treatment and a more elaborate measure of insight. Cerebral ventricular enlargement and tests of frontal lobe function did not correlate with insight, but there was a curious, strong association with left-handedness at both assessment points. Initial insight correlated significantly but weakly with insight at follow-up.

The assessment of insight in psychosis has concurrent validity and is a distinct aspect of psychotic phenomenology. It may, in part, have a neuropsychological basis.

We tested whether cortical and subcortical regional cerebral blood flow (rCBF) differs between patients with obsessive-compulsive disorder (OCD) and healthy controls. We then explored the relationship between rCBF and OCD mental state.

Thirty out-patients from the Maudsley Hospital with OCD as defined in DSM–III–R were scanned at rest using brain-dedicated, high-resolution, single photon emission tomography. RCBF was measured as uptake of 99mTc-HMPAO in 15 regions of interest and compared with rCBF data in 30 healthy people matched for age, sex and handedness. Symptom ratings were obtained using standard measures on the scanning day. Principal components factor analysis identified two distinct clinical dimensions: obsessive–compulsive (OC) and anxious–avoidant (AA). These were correlated with patients' rCBF measurements, using Spearman's rank correlation coefficient, and multiple regression coefficients calculated.

We found significant reductions in rCBF measurements of OCD patients compared with resting, healthy controls (F = 1.92, P = 0.04) in seven brain regions: the right and left superior frontal cortex, right inferior frontal cortex, left temporal cortex, left parietal cortex, right caudate nucleus and right thalamus. Regional differences were not secondary to generalised reduction in patients' brain perfusion. Reduced blood flow to the right inferior frontal cortex correlated significantly with illness severity (r = 0.37, P = 0.02). There was no relationship with age, age-of-onset, sex, handedness, depression or medication status. OC clinical dimension, concerning obsessions, compulsions and low mood, was significantly negatively correlated with left inferior frontal, medial frontal and right parietal rCBF. AA dimension, concerning anxiety and avoidance, was significantly positively associated with left and right superior frontal, right inferior frontal, medial frontal cortical, and right and left caudate and thalamic rCBF.

rCBF differs significantly between resting OCD patients and healthy controls, and separate clinical dimensions are associated with functionally distinct rCBF patterns.

This multicentre study compared a 16-week buspirone treatment with placebo in patients presenting with panic disorder with agoraphobia and also receiving cognitive behaviour therapy (CBT).

Double-blind testing was maintained until week 68, but not tested; 91 patients were included; 14 placebo-responders excluded; 77 patients randomised; 48 reached week 16 and 41 reached week 68.

At week 16, within-group analysis showed significant improvements in agoraphobia, panic attacks, and depression in both groups. Generalised anxiety improved only in CBT + buspirone. Between-group comparisons showed buspirone to have an effect on generalised anxiety and agoraphobia. Changes in degree of agoraphobia and depression were correlated in subjects on CBT + buspirone only. A significantly higher proportion of women, and of subjects showing high avoidance dropped out. Positive expectations regarding medication predicted success in both groups. At week 68, improvement was retained without significant buspirone effect.

Buspirone enhanced the effects of cognitive behaviour therapy on generalised anxiety and agoraphobia in the short term.

Age-associated memory impairment (AAMI) describes a non-disease ageing-related decline in memory. Pharmacological treatment trials have been reported and DSM–IV has introduced a term for the disorder. No prevalence study with the original criteria has previously been published.

An age-stratified sample was taken of 50–95-year-olds registered with a local health centre. Diagnosis of AAMI was made by questionnaire, cognitive testing, and medical and psychiatric assessment.

Prevalence rates for the total population and for the over-50s were estimated to be 5.8% and 18.5%, respectively. These rates are dramatically affected by minor alterations to individual criteria.

It appears AAMI is less common than previous estimates suggested, although the diagnostic criteria do not satisfactorily define people with an age-related decline in memory. Complaint of memory decline is more strongly correlated with measures of affect and personality than with measures of current memory test performance or estimates of memory decline.

Depression in later life is often thought to differ from that at other times of adulthood. The evidence for this is controversial but is important to any proposed organic model of depression in the elderly. Here, early- and late-onset depressions in later life are compared.

Fifty-seven depressed patients with a mean age of 74 were studied, 21 with an early onset (aged 59 or less) and 36 with a late onset. All were suffering from major depression according to DSM–III–R. The measures at entry included severity and symptoms, cognitive function, antecedent life events, physical health and vascular risk factors and/or vascular disease. We also recorded any family history of mood disorders, as well as the course of illness.

The anxiety item scores of the Hamilton Depression Rating Scale were significantly higher in those with an early onset, but otherwise symptoms differed little. Heritability was greater in the early-onset group. There was a striking association of vascular disease and/or risk with late-onset patients.

Vascular disease is associated with late-onset depression. This is consistent with the hypothesis that depression in later life is a more ‘biological’ disorder.

We examine the effects of cognitive-behavioural therapy (CBT) as an adjuvant to acute physical treatment and lithium maintenance therapy in reducing depression severity over a follow-up year in elderly depressed patients.

The study consists of three phases. During the acute treatment and continuation phase, 17 of 31 patients received CBT as an adjuvant to treatment as usual. During the maintenance phase of 1 year, subjects were entered into a double-blind, placebo-controlled study of low-dose lithium therapy.

Receiving adjuvant CBT significantly reduced patients' scores on the Hamilton Rating Scale for Depression during the follow-up year (repeated measures analyses of variance; P = 0.007). No significant differences were found between lithium and placebo maintenance therapy.

CBT can be adapted as an adjuvant therapy in the treatment of severely depressed elderly patients and reduces depression severity during follow-up. The prophylactic failure of long-term lithium therapy may be explained through poor compliance.

Pre-morbid intelligence level is routinely assessed in Alzheimer's disease using the National Adult Reading Test (NART). This practice is based on the assumption that pronunciation of irregular words remains unaffected by the disease process. Recent reports have suggested that reading ability may become compromised in moderately demented subjects.

Sixty-eight probable Alzheimer patients were classified into stages of severity (minimal, mild and moderate) using the Mini-Mental State Examination (MMSE). NART and demographic equations were used to estimate pre-morbid ability.

A significant correlation emerged between dementia severity and reading ability, NART v. MMSE scores, r = 0.46, P < 0.01. When the total sample was subdivided into moderate, mild and minimal subgroups, significant between-group differences emerged, despite the groups being well matched for age, sex, and years of full-time education. Pre-morbid IQ, as estimated by demographic regression equations, did not correlate with dementia severity.

NART performance is compromised in moderate Alzheimer disease, and the measure provides a serious underestimate of pre-morbid IQ in patients with an MMSE of 13 or less.

This paper describes a study of non-cognitive features of dementia in a population of 105 people over 50 years of age, with learning disability.

The study involved psychiatric assessment using the Psychiatric Assessment Schedule for Adults with a Developmental Disability (PAS-ADD), a semi-structured psychiatric interview developed specifically for people who have learning disability. Dementia was diagnosed using a combination of informant interviewing and observation of cognitive change over a three year period.

Sample members with definite dementia had higher levels of sleep difficulty, hypersomnia, irritability, inefficient thought, loss of interest and anhedonia. Discriminant function analysis showed that non-cognitive features could help discriminate between definite and probable cases.

Results support the previously reported observation for the general population that cognitive and non-cognitive features are only weakly related. The study suggests that psychiatric symptom information may be useful in screening for dementia in people with learning disability.

Morbid jealousy is an incompletely understood syndrome. The association between morbid jealousy and alcoholism is unclear.

Morbid jealousy was assessed in 207 male patients with alcohol dependence (DSM–III–R) using a semi-structured interview schedule administered both to patients and their spouses.

Thirty-four per cent (71 out of 207) suffered from morbid jealousy. Morbid jealousy had different manifestations. Some patients expressed it only when intoxicated, others even when sober and in some the jealousy took the form of a delusional disorder.

Alcoholism appears to have an aetiological role in the development of morbid jealousy. Detecting morbid jealousy when it is expressed only under the influence of alcohol has preventive implications.

Difficulties in family functioning have been noted since early descriptions of anorexia nervosa and may be of importance aetiologically. Previous studies have a number of methodological problems.

Thirty-five anorexic adolescents were age/sex matched with psychiatric and community controls. A diagnostic interview and a questionnaire, the Family Assessment Device (FAD) were administered to control subjects and their mothers. Anorexic families only received the McMaster Structured Interview of Family Functioning.

Multivariate analyses of FAD scores showed pathological ratings for psychiatric control but not anorexic families, compared with community controls. By contrast objective ratings revealed marked dysfunction in anorexic families (greater in the purging subgroup).

Family functioning in anorexic families is normal by self-report but not by an objective measure. Anorexic families in the purging subgroup appear most dysfunctional.