Generalised Anxiety Disorder Health Article

Many systematic reviews of randomised placebo-controlled trials in generalised anxiety disorder have been done since 1980, and in general these suggest that many drugs are effective in the short term and sometimes the medium term (table 2 ). Despite initial claims that panic disorder was a specific indication for antidepressant therapy but that generalised anxiety disorder was not responsive, ⁶⁹ evidence that antidepressants are effective treatments in both generalised anxiety and in the closely related panic disorder has now been shown. This dual efficacy was first shown by Kahn and colleagues ⁷⁰ with imipramine, and similar benefit has been shown with other tricyclic antidepressants. ^71,72 Most of these antidepressants were available before generalised anxiety disorder was a recognised diagnosis so few randomised controlled trials have been done for these treatments.

Table 2 shows many drugs of different classes that are similarly efficacious treatments for generalised anxiety disorder. ^25,72–76 Differences in efficacy have been shown in a few trials, but often the differences can be explained in terms of dosage, and these differences diminish after comparison of many meta-analyses. ⁷⁴

The choice of drug depends on the perceived nature of adverse effects and risks, the presence of co-existent depressive symptoms, and the need for an early onset-of-action. Benzodiazepines can be effective within 15–60 min, whereas buspirone takes up to 72 h to act and its action is often preceded by mild dysphoria. ⁷⁷ Antidepressants take up to 4 weeks to show effectiveness but clinically significant improvement can be noted after as little as 2 weeks. ^71,78 Benzodiazepines are associated with risk of dependence after long-term use ^77,79 and are more likely to lead to requests for long-term prescription than antidepressants (although there can sometimes be withdrawal effects with antidepressants ⁸⁰). The size of effect of treatment in generalised anxiety is greater for benzodiazepines than for antidepressants in the first 2 weeks of treatment, but at varying times from 3–8 weeks onwards the antidepressants show better effectiveness. ⁷¹ Initial treatment should probably be with combined benzodiazepines and an antidepressant and the benzodiazepine dose should be tapered off after 2–3 weeks when the antidepressant becomes effective. This method of treatment is now common in clinical practice ⁸¹ and is recommended frequently, ⁸² although it has not been tested in controlled trials.

Small differences between antidepressants have been seen in some studies, ^83,84 but are probably not enough to lead to firm conclusions regarding the superiority of any one class or individual drug over another. Most consensus guidelines and reviews suggest that tricyclic, SSRI, and serotonin and noradrenaline reuptake inhibitor (SNRI) classes have similar overall efficacy. ^25,74,85 The choice of treatment is determined by the range of adverse effects rather than by anxiolytic supremacy. Thus, if insomnia and restlessness are major symptoms it might sometimes be more appropriate to choose a tricyclic drug with sedative activity, such as amitriptyline or trimipramine rather than an SSRI or SNRI because these drugs can cause an increase in anxiety and restlessness in the early stages of treatment. The range of adverse effects with SSRIs (nausea, dizziness, anorexia) is generally considered to be less serious than that of tricyclic antidepressants (dry mouth, sedation, postural hypotension, difficulty in micturition), so an SSRI (eg, escitalopram or paroxetine) is usually recommended as first-line treatment. ^25,72,85 Escitalopram 10 mg per day seemed superior to paroxetine 20mg per day in a 12-week randomised placebo-controlled trial ⁸⁶ and was better tolerated in a comparative 6-month study. ⁸⁷ The SNRI venlafaxine is effective ²⁵ but because of tolerability concerns is often only recommended for use as a second line treatment. ⁸⁵

Generalised anxiety disorder is a chronic condition, so long-term treatment should be anticipated and planned accordingly. Most evidence of efficacy is from acute treatment studies and recommendations for treatment beyond 12 weeks are on the basis of few data. The SSRIs escitalopram and paroxetine have both shown efficacy in placebo-controlled relapse-prevention studies. ^88,89 Continued treatment with venlafaxine is better than continued treatment with placebo, in the proportion of patients that achieve symptomatic remission, but venlafaxine was not efficacious at relapse prevention. ⁹⁰ Relapse is frequent after drugs are withdrawn, even after the period of withdrawal symptoms has passed. Randomised trials including both treatment and withdrawal components have shown that the effectiveness of placebo increases strikingly. ⁹¹ The concern about benzodiazepines causing dependence has led to recommendations that these drugs should be avoided in the long-term, ^25,72,73,85 but the alternative of continuous antidepressant therapy is supported by little evidence. Buspirone is an anxiolytic drug that has been suggested to be of value in the long-term treatment of generalised anxiety disorder because it does not result in addiction. Buspirone is thought to be especially useful in patients with alcohol dependence, but again this use has little supporting evidence. ⁹²

Other drugs might also be of use in the treatment of generalised anxiety disorder. β blockers such as propranolol, although widely used for anxiety in primary care, ⁹³ have unproven efficacy ⁹⁴ in generalised anxiety disorder, although studies done before the disorder was delineated suggested some value in patients who had marked somatic manifestations of anxiety that seemed to be mediated through β-adrenergic receptors. ⁹⁵ Some evidence suggests that propranolol combined with a benzodiazepine could be more effective than a benzodiazepine alone in generalised anxiety disorder ⁹⁶ and this combination could help with the subsequent withdrawal of the benzodiazepine. ⁹⁷

Monoamine oxidase inhibitors such as phenelzine might also be effective in the treatment of intractable generalised anxiety disorder. This class of drugs is highly effective in severe anxiety but their potential for adverse reactions with tyramine-containing foods (mainly cheese) and sympathomimetic drugs has restricted their use and they have not been formally tested in generalised anxiety disorder. However, their effectiveness in so-called endogenous anxiety ⁹⁸ suggests a role for these drugs in secondary care settings when other treatments have failed. The safer reversible monoamine oxidase inhibitor, moclobemide, is not as effective as the irreversible monoamine oxidase inhibitors ⁹⁹ but has shown efficacy in social phobia.

The psychological symptoms of anxiety might respond better to antidepressant drugs than to benzodiazepines, but few comparator-controlled studies have been done, and most showed no significant differences in efficacy between active compounds. ^74,75 Antidepressant treatment is preferable to benzodiazepine anxiolytics in cothymia and other mixed conditions, because benzodiazepines have restricted efficacy against depressive symptoms. ^25,73,74 Advice that benzodiazepines should only be prescribed for up to 4 weeks in regular doses and then tapered off has provoked some reaction since this method can be less efficacious than placebo. ⁹¹ Some clinicians argue that patients who respond well to benzodiazepines have chronic disorders for whom continued treatment is desirable and justifiable, ¹⁰⁰ but this remains a controversial view.

Although low doses of antipsychotic drugs have been used in primary care settings to treat anxiety symptoms, formal evidence of efficacy of antipsychotic treatment as monotherapy seems limited to a single placebo-controlled study of the conventional antipsychotic, trifluoperazine: although it was superior to placebo from the first week, it was poorly tolerated. ¹⁰¹ The sedative effects of antihistamines have sometimes been used to dampen troublesome anxiety symptoms, and this strategy is supported by the proven efficacy of hydroxyzine as a short-term treatment. ^102,103

Despite the abundance of guidelines, there is confusion over the best method of treating generalised anxiety disorder with drugs. In practice, patients who were formerly treated with benzodiazepines alone are now being treated with benzodiazepines and antidepressants (usually SSRIs) ¹⁰⁴ with the intention of stopping the benzodiazepine treatment. However, removal of the benzodiazepine is not always possible.

Little is known about the management of patients who have not responded to first-line treatment. An increase in overall response with dose-escalation in patients after an initial non-response to a lower dose has not been proven, but switching between treatments with proven efficacy could be helpful. ²⁵ Small placebo-controlled studies with the antipsychotic drugs olanzapine and risperidone suggested they can enhance the response to the SSRI fluoxetine ¹⁰⁵ or to anxiolytics. ¹⁰⁶