The past two decades have witnessed an increasing interest in research on anxiety syndromes and disorders. Significant advances in basic and clinical research, accompanied by animated ongoing discussions about the most appropriate, reliable and valid way to classify and assess anxiety disorders, have changed our understanding of anxiety disorders and affected (or some might say confused) our research and treatment practice.

Background Mental disorders impose a multi-billion dollar burden on the economy each year; translating the burden into economic terms is important to facilitate formulating policies about the use of resources.

Method For direct costs, data were obtained from national household interview and provider surveys; for morbidity costs, a timing model was used that measures the lifetime effect on current income of individuals with mental disorders, taking into account the timing of onset and the duration of these disorders, based on regression analysis of Epidemiologic Catchment Area study data.

Results The total economic costs of mental disorders amounted to US$ 147.8 billion in 1990. Anxiety disorders are the most costly, amounting to $46.6 billion, or 31.5% of the total; schizophrenic disorders accounted for $32.5 billion, affective disorders for $30.4 billion, and other mental disorders for $38.4 billion.

Conclusions Mental illnesses, especially anxiety disorders, are costly to society. Although anxiety disorders have a higher prevalence than affective disorders and schizophrenia, use of medical care services is lowest for anxiety disorders. Anxiety disorders appear to be under-recognised and untreated even though treatment interventions have been shown to be effective and can be delivered in a cost-efficient manner.

Background The paper describes the rationale, sensitivity and specificity of the Anxiety Screening Questionnaire (ASQ), a disorder-specific screening instrument for use in primary care.

Method Two hundred and fifty subjects sampled from psychiatric, primary care settings and the community, participated in a test-retest reliability as well as a procedural validity study, using the M–CIDI with DSM–IV algorithms as a diagnostic yardstick.

Results The ASQ was found to be easy to administer and acceptable and efficient in terms of sensitivity and specificity for generalised anxiety syndromes. The test– retest item reliability was good to excellent with kappa values of 0.6 or above. As compared with the validity standard, the DSM – IV/CIDI diagnoses caseness sensitivity was generally high (above 82%) for all diagnostic domains covered, whereas the specificity was only high for DSM – IV threshold and subthreshold generalised anxiety disorder.

Conclusions These preliminary findings demonstrate the usefulness of this anxiety screening questionnaire, constructed closely following the guidelines of specific diagnostic criteria.

Background This study explored the prevalence, socio-demographic characteristics and severity of different anxiety syndromes in five European primary care settings, as well as medical help-seeking, recognition by general practitioners (GPs) and treatment prescribed.

Method The data were collected as part of the WHO study on Psychological Problems in General Health Care. Among 9714 consecutive primary care patients, 1973 were interviewed using the Composite International Diagnostic Interview. Reason for contact, ICD–10 diagnoses, severity and disability were assessed. Recognition rates and treatment prescribed were obtained from the GPs.

Results Anxiety syndromes, whether corresponding to well-defined disorders or to subthreshold conditions, are frequent in primary care and are associated with a clinically significant degree of severity and substantial psychosocial disability. Their recognition by GPs as well as the proportion treated are low.

Conclusions Since people with subthreshold anxiety show a substantial degree of disability and suffering, GPs may consider diagnostic criteria to be insufficient. However, their awareness of specific definitions and treatment patterns for anxiety disorders still needs a lot of improvement both for patients' well-being and for the cost resulting from non-treatment.

Background The co-occurrence of anxiety disorders with other mental, addictive, and physical disorders has important implications for treatment and for prediction of clinical course and associated morbidity.

Method Cross-sectional and prospective data on 20 291 individuals from the Epidemiologic Catchment Area (ECA) study were analysed to determine one-month, current disorders, one-year incidence, and one-year and lifetime prevalence of anxiety, mood, and addictive disorders, and to identify the onset and offset of disorders within the one-year prospective period.

Results Nearly half (47.2%) of those meeting lifetime criteria for major depression also have met criteria for a comorbid anxiety disorder. The average age of onset of any lifetime anxiety disorder (16.4 years) and social phobia (11.6 years) among those with major depression was much younger than the onset age for major depression (23.2 years) and panic disorder.

Conclusions Anxiety disorders, especially social and simple phobias, appear to have an early onset in adolescence with potentially severe consequences, predisposing those affected to greater vulnerability to major depression and addictive disorders.

Background This study explores the temporal relationship between anxiety and major depressive disorders in a cohort of patients with current major depression.

Method Current prevalence and lifetime history of specific anxiety disorders were assessed using the Structured Clinical Interview for DSM–III–R Diagnosis (SCID–P) in 85 patients with DSM–III–R major depression. Consensus DSM–III–R diagnoses were assigned by at least two psychiatrists or psychologists.

Results Twenty-nine per cent met criteria for at least one current anxiety disorder and 34% had at least one anxiety disorder at some point in their lives. The mean (s.d.) age of onset of anxiety disorder in the depressed patients with comorbid social or simple phobia (15 (9) years) was significantly younger than was that of their major depression (25 (9) years). In contrast, the mean (s.d.) age of onset of anxiety in patients with comorbid panic or OCD (20 (8) years) was similar to that seen for their major depression (21 (9) years). In patients with major depression with comorbid anxiety disorders, both the social phobia (10 of 13) and simple phobia (4 of 4) were more commonly reported to start at least two years prior to their major depression in contrast to depressives with comorbid panic (3 of 10 subjects) – Fisher's exact test, P=0.01.

Conclusions Early-onset social and simple phobias appear to be risk factors for later onset of major depression.

Background Although there have been many changes in the diagnosis of anxiety and depressive disorders in the past 20 years there have been few comparative enquiries into the clinical outcome of greater diagnostic categories. We therefore compared the outcome of all studies which compared the outcome of specific anxiety and depressive disorders using the standard procedures of systematic review.

Method A Medline search was carried out of all studies comparing the outcome of anxiety and depressive disorders or mixed anxiety-depressive disorders in which information was available separately for each disorder.

Results Eight studies satisfied the search criteria (all involving a period of observation of two years or greater); only one of these included randomisation of treatment and comparison between specific anxiety disorder outcome. There was a somewhat better outcome in patients with depressive disorders compared with anxiety ones, and strong evidence that both anxiety and depressive disorders singly had better outcomes than comorbid mixed disorders.

Conclusion Comorbid anxiety-depressive disorders have a poor outcome compared with single anxiety and depressive disorders, and there is some evidence that anxiety disorders have a worse outcome than depressive ones.

Background Pharmacological and psychological treatments for anxiety are often combined in clinical practice but there is little research from which to predict the effects.

Method The theoretical outcomes of combining treatments and methods of investigating these as well as methodological difficulties are described. Studies which have been completed in anxiety disorders are reviewed. A double-blind trial, using a factorial design, evaluated buspirone v. placebo and anxiety management training v. nondirective therapy in 60 patients with generalised anxiety disorder (GAD).

Results Relatively few germane studies have been carried out in the anxiety disorders except for panic disorder with agoraphobia. There is some evidence that short-term, combined treatment does confer additional benefits which are evident both in speed of onset and lasting remission. All four treatment combinations proved effective in the short-term treatment of GAD.

Conclusions More studies examining combined treatment are needed. Although differences may not be apparent at the end of the treatment period, psychological treatment appears to confer advantages at follow-up.

Background DSM–IV generalised anxiety disorder (GAD) has a high lifetime prevalence, but subthreshold anxiety states are even more common, particularly in family practice.

Method Generalised anxiety is conceptualised as a spectrum of disorders, with transient anxiety at one end and GAD at the other.

Results Based on longterm experience with family practice patients, the authors suggest that most anxious patients, wherever on this continuum they are placed, could be treated with short-term, possibly intermittent, rather than chronic anxiolytic therapy. Data are presented which show that 50% of chronic GAD patients are only in need of such short-term intermittent therapy.

Conclusions Further clinical research is needed to refine short-term, intermittent treatments for anxiety spectrum disorders, to make effective treatments available to those suffering from anxiety but falling short of diagnostic criteria for GAD, and to target more effectively the different treatment strategies.

Background Abecarnil, a novel anxiolytic beta-carboline, was investigated in five four-week double-blind, European multicentre studies. Overall, 451 patients with generalised anxiety disorder were randomised to abecarnil, 461 to placebo and 464 to active controls.

Method Data includes inferential statistics based on individual studies and descriptive analysis of 323 patients in open-label abecarnil longterm continuation up to 52 weeks.

Results Abecarnil was safe, the most frequent adverse event being drowsiness. Onset of effect was at week I. At week 4 the Hamilton Anxiety Scale score had improved by 12–13 points on average. Due to notably large and variable placebo effects abecarnil was not consistently superior to placebo. No rebound or withdrawal symptoms were observed after fast-tapered discontinuation. Safety, extent of efficacy and incidence of rebound or withdrawal did not change during longterm treatment.

Conclusions Abecarnil is safe and effective. Further research into its therapeutic potential seems warranted.