Serotonin (5-HT) interacts with multiple brain 5-HT receptor subtypes to influence a wide range of behaviours. Three main families of 5-HT receptors (5-HT1, 5-HT2 and 5-HT3) have been described which differ in their binding affinity for selective ligands, their receptor-effector coupling mechanisms, and the behavioural processes they regulate. Nevertheless, manipulation of several different 5-HT receptor subtypes (5-HT1A, 5-HT1c, 5-HT2 and 5-HT3) may produce anxiolytic effects; 5-HT1A and 5-HT2 receptors may be involved in the aetiology of major depression and the therapeutic effects of antidepressant treatment; and 5-HT3 receptors have been linked to reward mechanisms and cognitive processes. These advances offer therapeutic possibilities, the value of which can only be satisfactorily assessed by controlled clinical trials.

The principles of measuring anxiety are reviewed. Scales of anxiety differ in their time focus: state-anxiety scales measure the present or recent past, whereas trait-anxiety scales measure the ways the subject generally feels. Commonly used scales have adequate validity, but there is incomplete agreement on the boundaries of the construct of anxiety. There are consistent differences in sensitivity among anxiety scales, and they vary greatly in the number of items, number of cues and in the method of scoring. For self-rating scales a large number of items with only two cues (e.g. ‘yes’ and ‘no’), or only two scores for each item, increases the sensitivity of the scale for screening as well as for detecting differences between drug and placebo. A large number of cues decreases the sensitivity of a self-rating scale; it is unknown whether a large number of cues is advantageous in group research, but a small number of items is adequate both for rating and self-rating scales.

The similarities and differences of comorbidity in treated and untreated samples with anxiety disorders were examined using data from the Munich Follow-up Study: 133 subjects with anxiety and depressive disorders and 101 former in-patients at the Max Planck Institute for Psychiatry. Diagnoses were based on the DIS, not using the optional DSM—III exclusion rules. In these epidemiological and clinical samples, 69% and 95% respectively had at least two diagnoses. The epidemiological sample was clearly differentiated from the clinical sample by age of onset. The development of both depressive episodes and substance disorders in the two samples was mostly secondary to the development of anxiety problems. The outcome for subjects with both anxiety and depressive disorders tended to be worse than that for those with anxiety alone, regardless of whether a depressive episode was present at the time of the follow-up investigation. The usefulness is underlined of the comorbidity concept based on operationalised diagnosis without the exclusion rules offered by DSM—III and DSM—III—R.

The augmentation of serotonin (5-HT) neurotransmission attenuates alcohol consumption, whereas depletion enhances use. A number of serotonin-specific pharmacological probes appear to be effective in reducing the voluntary consumption of alcohol. The 5-HT1A receptor agonist buspirone is an anxiolytic which has been shown to diminish the desire to consume alcohol in anxious alcoholic patients. Thus, serotonin may represent a common denominator for a spectrum of behavioural disorders including anxiety and alcoholism. Novel serotonin drugs, such as the azapirones, may usefully supplement conventional treatment strategies for substance abuse.

Buspirone is an anxiolytic drug from the azapirone family of molecules. It differs chemically and pharmacologically from the benzodiazepines. Although its profile of efficacy is comparable with that of benzodiazepines, it produces less drowsiness, less psychomotor impairment, less alcohol potentiation, and has less potential for addiction or abuse. Buspirone also appears to have efficacy in major depressive disorders, in comparison with placebo, but its activity in panic disorders is less impressive. It may diminish alcohol dependence both in animals and in chronic alcoholics. Clinical studies in the elderly show no important difference from younger patients in safety and efficacy profile, pharmacokinetics, and dosage requirement. The drug appears to be well tolerated in primary care settings and to be free of adverse clinical interactions with many drugs that might be used concomitantly. However, because its pharmacology differs from that of conventional anxiolytics, patients need to be informed about both its gradual onset of action and absence of euphoria and immediate sedation.

Buspirone is a newer anxiolytic which differs chemically and pharmacologically from the benzodiazepines. The question as to whether buspirone can induce rebound, dependence, or abuse can be examined through a variety of clinical and control studies, as well as animal studies, together with meticulous follow-up of adverse reports. It is concluded that, on present evidence, the short-term use of buspirone is unlikely to be followed by rebound or long-term use through dependence; nor is abuse likely to be a problem.