• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Mental Health (UK). Generalised Anxiety Disorder in Adults: Management in Primary, Secondary and Community Care. Leicester (UK): British Psychological Society; 2011. (NICE Clinical Guidelines, No. 113.)



This guideline is concerned with the treatment and management of adults with a diagnosis of GAD in primary and secondary care. GAD is one of a range of anxiety disorders including panic disorder (with and without agoraphobia), PTSD, OCD, social phobia, specific phobias (for example, of spiders) and acute stress disorder.

GAD commonly coexists with other anxiety disorders and with depressive disorders, as well as a variety of physical health disorders. ‘Pure’ GAD in the absence of another anxiety or depressive disorder is less typical than comorbid GAD. This guideline is relevant to both people with pure and comorbid GAD. The NICE guideline on case identification and referral for common mental health disorders will provide further guidance on identification (NICE, 2011).


2.2.1. Symptoms, presentation and patterns of illness

Anxiety is a prominent symptom of many psychiatric disorders but it is only comparatively recently that several distinct anxiety disorders have been recognised in classificatory systems. The key feature of GAD is worry and apprehension that is out of proportion to the circumstances. The worries are typically widespread, involve everyday issues and have a shifting focus of concern. The affected person finds the worries difficult to control, and this can result in decreased occupational and social functioning (Tyrer & Baldwin, 2006; Bitran et al., 2009).

As well as worry that is excessive, generalised and difficult to control, people with GAD experience other psychological and somatic symptoms of anxiety. Psychological symptoms include irritability, poor concentration, increased sensitivity to noise and sleep disturbance, typically difficulty falling asleep. Somatic symptoms of GAD can manifest in many different ways. For example, an overactive autonomic nervous system can lead to sweating, dry mouth, palpitations, urinary frequency, epigastric discomfort and frequent and/or loose bowel motions, while hyperventilation may result in feelings of shortness of breath and dizziness. Increased muscle tension is a common accompaniment of persistent anxiety and may be experienced as restlessness, inability to relax, headaches and aching pains, particularly in the shoulders and back (Gelder et al., 2006).

GAD is frequently comorbid with other mental disorders, which can complicate its presentation. The rates of comorbidity vary between studies with estimates of between 68 and 93% of comorbidity with another axis 1 mental health disorder (Carter et al., 2001; Hunt et al., 2002; ESEMeD/MHEDEA 2000 Investigators, 2004). Comorbid disorders that are particularly common include depressive disorders (specifically major depression and dysthymia), other anxiety disorders (especially panic disorder, social phobia and specific phobias) and somatoform disorders (Bitran et al., 2009; Carter et al., 2001; Hunt et al., 2002; Grant et al., 2005; Kessler et al., 2005b). There is also significant comorbidity with substance misuse especially among men (Grant et al., 2005; Kessler et al., 2005b).

GAD also often co-occurs with physical health problems such as arthritis and gastrointestinal and respiratory disorders and may mimic the presentation of some physical conditions (for example, hyperthyroidism) (Culpepper, 2009; Roy-Byrne et al., 2008; Sareen et al., 2006). Due to the somatic symptoms of anxiety, which are central to GAD, and physical comorbidities, people with GAD who present in primary care may emphasise somatic problems or sleep disturbance, rather than excessive worry or psychological symptoms of anxiety (Rickels & Rynn, 2001).

2.2.2. Course and prognosis

Most clinical studies suggest that GAD is typically a chronic condition with low rates of remission over the short and medium-term. Evaluation of prognosis is complicated by the frequent comorbidity with other anxiety disorders and depression, which worsen the long-term outcome and accompanying burden of disability (Tyrer & Baldwin, 2006). In the Harvard-Brown Anxiety Research Program, which recruited participants from Boston hospitals, the mean age of onset of GAD was 21 years, although many participants had been unwell since their teens. The average duration of illness in this group was about 20 years and despite treatment the outcome over the next 3 years was relatively poor, with only one in four showing symptomatic remission from GAD (Yonkers et al., 1996). The proportion of people who became free from all psychiatric symptomatology was smaller, about one in six. In people who remitted from GAD the risk of relapse over the next year was about 15%, increasing to about 30% in those who achieved only partial symptomatic remission (Yonkers et al., 1996).

The participants in the above study were recruited from hospital services and may not be representative of GAD in general. In a naturalistic study in the UK, Tyrer and colleagues (2004) followed up people with anxiety and depression identified in psychiatric clinics in primary care and found that 12 years later 40% of those initially diagnosed with GAD had recovered, in the sense that they no longer met criteria for any Diagnostic and Statistical Manual of Mental Disorders 3rd edition (DSM-III; American Psychiatric Association [APA], 1980) psychiatric disorder. The remaining participants remained symptomatic, but GAD was still the principal diagnosis in only 3% of trial participants; in the vast majority conditions such as dysthymia, major depression and agoraphobia were now more prominent. This study confirms the chronic and fluctuating symptomatic course of GAD in clinically-identified people. It should be noted, however, that the majority of people with GAD in the community do not seek medical help for their symptoms (Wittchen & Jacobi, 2005), and the course of the illness in these circumstances is not established.

2.2.3. Disability and mortality

As is the case with major depression, GAD is associated with a substantial burden of disability, equivalent to that of other chronic conditions such as arthritis and diabetes (Wittchen, 2002). Outcome studies suggest that anxiety disorders are more chronic than other common mental disorders (Tyrer et al., 2004) and there is evidence that comorbid depression and anxiety has a worse prognosis, with more associated disability and more persistent symptoms than either depression or anxiety disorders alone (Kroenke et al., 2007). There is also evidence in the community that anxiety disorders are independently associated with several physical conditions, and this comorbidity is significantly associated with poor quality of life and disability (Sareen et al., 2006). This morbidity comes with high associated health and social costs (Simon et al., 1995).

Studies have shown that the presence of GAD is also associated with significant impairments in occupational and social functioning. For example, over 30% of people with GAD showed an annual reduction of work productivity of 10% or more compared with 8% of people with major depression. The figure for people with comorbid GAD and depression was over 45% (Wittchen et al., 2000). A large part of the economic cost of anxiety disorders is attributable to the costs of non-medical psychiatric treatment. People with GAD have increased numbers of visits not only to primary care doctors but also to hospital specialists, particularly gastroenterologists (Kennedy & Schwab, 1997; Wittchen, 2002). This may be a consequence of the distressing somatic symptoms which many people with GAD experience.

GAD also carries a considerable cost in personal suffering – in the Harvard-Brown Anxiety Research Program noted above, one third of people had never married and unemployment was higher than average (Yonkers et al., 1996). Suicidal ideation and suicide attempts are significantly increased in GAD compared with the general population, particularly in women, and this increase is still greater in the presence of comorbid major depression (Cougle et al., 2009).

2.2.4. Incidence and prevalence

The estimated proportion of people in England with GAD was 4.4% in the most recent Adult Psychiatric Morbidity in England survey (McManus et al., 2009), a figure that has varied little across the three survey years 1993, 1997 and 2007. This figure is at the upper end of estimates of point and annual prevalence of 2.1 to 4.4% in English speaking countries (Grant et al., 2005; Hunt et al., 2002; Kessler & Wang, 2008) with lower rates of 0.8 to 2.2% reported from other European countries (Lieb et al., 2005; Wittchen & Jacobi 2005). Worldwide estimates of the proportion of people who are likely to experience GAD in their lifetime vary between 0.8% and 6.4% (Lieb et al., 2005; Grant et al., 2005; Kessler & Wang, 2008).

Prevalence rates have generally been found to be between 1.5 and 2.5 times higher in women than men. In the Adult Psychiatric Morbidity in England survey (McManus et al., 2009), the rates were 3.4% for men and 5.3% for women. In terms of age, epidemiological studies have generally found GAD to be less common in older age groups (over 55 years) although there are some exceptions. Some studies have also found GAD to be less common in younger adults (younger than 35 years).

Evidence from the US on ethnicity and race differences in GAD rates is inconsistent, with studies finding increased (Blazer et al., 1991), decreased (Grant et al., 2005) and no difference (Wittchen et al., 1994) in rates between white and one or more of black, Asian and Hispanic groups. Numbers of minority ethnic groups sampled in the Adult Psychiatric Morbidity in England survey (McManus et al., 2009) were too small to draw conclusions about possible differences, although proportions of the black and South Asian groups with GAD in the sample (both male and female) were higher than the equivalent proportions for white interviewees.

Socioeconomic factors associated with GAD are lower household income (Grant et al., 2005; McManus et al., 2009), lack of tertiary qualifications (Hunt et al., 2002) and unemployment (Hunt et al., 2002). Divorce, separation and death of a partner are also associated with an increased likelihood of GAD.

2.2.5. Diagnosis

Diagnostic criteria and methods of classification of anxiety disorders have changed substantially over the years. Historically what we now consider to be GAD was subsumed under ‘anxiety neurosis’. It first appeared as a separate diagnosis in 1980 with the introduction of DSM-III (APA, 1980). In DSM-III it was a residual category to be used only when an anxiety disorder could not be classified under another diagnosis. It was only with the DSM-III revision in 1987 (DSM-III-R; APA, 1987) that it became a well defined condition in its own right. DSM-III-R also changed the DSM-III minimum duration requirement from 1 month to 6 months and introduced excessive worry as a central feature. Some of the developments in DSM-III-R were later reflected in the International Classification of Diseases – the Classification of Mental and Behavioural Disorders 10th revision (ICD-10; World Health Organization [WHO], 1992), although without the same focus on worry. The introduction of DSM-IV in 1994 (APA, 1994) further streamlined and refined the criteria, in particular focusing less on somatic symptoms of anxiety and replacing the DSM-III-R criterion that the worry is ‘unrealistic’ with a criterion that the worry is ‘difficult to control’.

DSM-IV and ICD-10 have overlapping but different diagnostic features for GAD. DSM-IV emphasises worry (‘apprehensive expectation’), including the feature that the worry is difficult to control, while ICD-10 focuses more on somatic symptoms of anxiety, particularly autonomic reactivity and tension. DSM-IV requires two major symptoms (6 months or more of excessive anxiety and worry, occurring on more days than not, about a number of events and activities and difficulty controlling the worry) and three or more additional symptoms from a list of six. ICD-10, as operationalised in the ICD-10 Diagnostic Criteria for Research (ICD-10-DCR; WHO, 1993), requires 6 months or more prominent tension, worry and feelings of apprehension, and four from a list of 22 symptoms, of which at least one must be from a list of four autonomic symptoms (palpitations, sweating, trembling, dry mouth).

In line with the previous guideline on GAD (NICE, 2004a) and other NICE guidelines on anxiety disorders and depression (NICE, 2005a, b; 2009b) the GDG used DSM-IV, rather than ICD-10 to define the diagnosis of GAD, because the evidence base for treatments nearly always uses DSM-IV.

As there is now greater recognition of the need to consider ‘subthreshold’ depression in terms of human and economic costs and the risk of future major depression (Rowe & Rapaport, 2006), there has also been recent attention given to subthreshold GAD. Relaxing the DSM-IV requirements of duration, excessive worry and/or three associated symptoms more than doubles the estimated prevalence of GAD (Ruscio et al., 2007). Cases of subthreshold GAD have similar but reduced comorbidities, with persistence, impairment and sociodemographic correlates all being significantly associated with an elevated risk of subsequent psychopathology (Kessler et al., 2005a; Ruscio et al., 2007). The implication is that, in clinical practice, identification of subthreshold GAD may be helpful for prevention of future disorder.


The aetiology of GAD is multifactorial and involves psychological, social and biological factors. Interpretation of experimental data is complicated by changes in diagnostic practice and the frequent occurrence of comorbidity, particularly with major depression (Yonkers et al., 1996). On the other hand, anxiety (or more precisely, fear) is readily modelled in animal experimental studies, and the brain circuitry relevant to fear has been characterised in both animals and humans (Engel et al., 2009). One influential formulation (‘the theory of triple vulnerability’) regards GAD as arising from three distinct kinds of vulnerability: a generalised biological, a generalised psychological and a specific psychological vulnerability (Barlow, 2000; Bitran et al., 2009).

Anxiety disorders run in families. For example, a family study found that the risk of GAD in first-degree relatives of people with GAD was five times that in control groups (Noyes et al., 1987), although specific genes conferring vulnerability to GAD have not yet been reliably identified. Indeed the genes involved in the transmission of GAD appear to increase susceptibility to other anxiety disorders such as panic disorder and agoraphobia as well as major depression (Kendler, 1996; Hettema et al., 2001; 2005). There is also genetic overlap between GAD and the temperamental trait of neuroticism, which is itself a predisposing factor for GAD (Hettema et al., 2004). Overall the findings suggest that genetic factors play a significant though moderate role in the aetiology of GAD, that these factors predispose people to a range of anxiety and depressive disorders rather than GAD specifically, and that environmental factors are important in determining the nature of the emotional disorder experienced by a particular person.

Several environmental factors are known to predispose individuals to GAD. These can act remotely or as contemporaneous triggers to the disorder. For example, good parenting experiences are important in providing children with a secure base from which to explore the world, and problems in child-parent attachment have been linked to feelings of diminished personal control of potentially threatening events (Barlow, 2000). Such feelings could plausibly contribute to the risk of experiencing anxiety disorders. Studies suggest that adults with GAD report experiencing parental styles characterised by overprotection and lack of emotional warmth (Silove et al., 1991). Similar findings have been reported in other anxiety disorders and depression (Parker et al., 1995), which suggest that certain parenting styles may act as a psychological vulnerability factor for a range of subsequent emotional disorders. Similar comments apply to other kinds of childhood adversity such as neglect, abuse, maternal depression and family disruption, which increase the risk of experiencing GAD in adulthood as well as other anxiety and depressive disorders (Brown & Harris, 1993; Halligan et al., 2007; Safren et al., 2002). More recent stressful life events are also known to be involved in the onset of emotional disorders including GAD (Roemer et al., 1996). A study by Kendler and colleagues (2003) showed that stressful life events characterised by loss increased the risk of both depression and GAD; however, life events characterised by ‘danger’ (where the full import of the event was yet to be realised) were more common in those who subsequently developed GAD.

Particular coping and cognitive styles also predispose individuals to the development of GAD, although it is not always easy to distinguish predisposition from the abnormal cognitions seen in the illness itself. As noted above, it is believed that people who lack a sense of control of events and personal effectiveness, perhaps through early life experiences, are more prone to anxiety disorders (Barlow, 2000). Such individuals may also demonstrate trait-like cognitive biases in the form of increased attention to potentially threatening stimuli, overestimation of environmental threat and enhanced memory of threatening material. This has been referred to as the ‘looming cognitive style’, which appears to be a general psychological vulnerability factor for a number of anxiety disorders (Reardon & Nathan, 2007). More recent cognitive formulations have focused on the process of worrying itself, which is of central importance in the diagnosis of GAD. Studies suggest that people at risk of GAD use worry as a positive coping strategy to deal with potential threats, whereby the person worries until they feel reassured that they have appraised all possible dangers and identified ways of dealing with them. However, this can lead to ‘worry about worry’, when individuals come to believe, for example, that worrying in this way, while necessary for them, is also uncontrollable and harmful. This ‘metacognitive belief’ may constitute a transitionary stage between excessive, but normal, worrying and GAD (Wells, 2005).

Studies of both animal and human subjects suggest that the amygdala plays a central role in the processing of information relevant to threat and fear (Le Doux, 2000). Activation of the amygdala can occur prior to conscious appreciation of threat but there are strong connections between the amygdala and areas of prefrontal cortex involved in the conscious experience and regulation of emotion (Le Doux, 2000; Phillips et al., 2003). Another structure involved in anxiety is the hippocampus, which is important in relating fearful memories to their environmental context (Fanselow, 2000). The hippocampus forms part of a ‘behavioural inhibition system’, which is activated by potential threats, and has the ability in these circumstances to suspend ongoing behaviours (Gray, 1982). Brain imaging studies of individuals with high trait anxiety and people with GAD have shown exaggerated responses in both the amygdala and prefrontal cortex during presentation of emotionally threatening stimuli (Bishop et al., 2004; Nitschke et al., 2009). It is therefore possible that pre-existing abnormalities in this circuitry might predispose people to GAD and other anxiety disorders.

The neural circuitry involved in fear and anxiety is modulated by brain neuro-transmitters and other chemical mediators including hormones (Dedovic et al., 2009). A relevant hormonal system is the hypothalamo-pituitary-adrenal axis (HPA), which regulates cortisol secretion. Adversity experienced in childhood and current stresses can alter the pattern of cortisol secretion in adult life, and there is an extensive literature on the role of HPA axis dysfunction in major depression (for example, Pariante & Lightman, 2008). HPA axis activity in people with GAD has been much less studied but there is some evidence that GAD, like depression, is associated with excessive glucocorticoid secretion (Mantella et al., 2008). The monoamine neurotransmitters, serotonin and noradrenaline, can alter fear processes in animals and have extensive inputs to the relevant neural circuitry, including the amygdala and the behavioural inhibition system (Bitran et al., 2009; Garner et al., 2009). In addition, selective serotonin reuptake inhibitors (SSRIs) are widely used in the treatment of GAD (Baldwin et al., 2005). Despite this there is only modest evidence that abnormalities in serotonin and noradrenaline are involved in the pathophysiology of GAD, though more work needs to be carried out with ligand neuroimaging to resolve this issue (Garner et al., 2009). In the same way, pharmacological manipulation of gamma-aminobutyric acid (GABA) neurones and their associated benzodiazepine receptors clearly have profound effects on the experience of fear and anxiety in animals and humans (Kalueff & Nutt, 2007) but again there is only modest evidence that abnormalities in GABA neurotransmission or benzodiazepine receptor function are involved in the aetiology of GAD (Garner et al., 2009).

Overall there is good evidence that both genetic factors and early life difficulties can predispose people to a range of emotional disorders, including GAD. More specific risk factors for GAD, presumably occurring in combination with these more generalised vulnerabilities, include certain kinds of life events and particular individual cognitive styles involving the use of worrying as a coping strategy. The neural circuitry involved in fear and anxiety has been well delineated in brain imaging studies and abnormalities in both people with GAD and non-clinical subjects with high trait anxiety have been described in relevant brain regions. It seems likely that these neural changes are associated with abnormal cognitions, such as increased attention to threat, that are seen in people with GAD and those at risk of the disorder. There is much knowledge on how particular neuropharmacological manipulations can influence anxiety. While this information has proved helpful in developing pharmacological treatment, the role of neurotransmitters and other chemical mediators in the aetiology of GAD is currently unclear.


2.4.1. Detection, recognition and referral in primary care

Relative to its prevalence in the community, GAD is more common in primary care occurring in about 5% of attendees, and is the most common anxiety disorder seen in this setting. A recent international review of some of the larger general population surveys reported 12-month prevalence rates of 5.6 to 18.1% for anxiety disorders, of which GAD and panic disorder together accounted for over half of the prevalence figures (Baumeister & Hartner, 2007).

General practitioner (GP) rates of diagnosis and treatment of anxiety disorders are much lower than expected from the prevalence figures (Wittchen & Jacobi, 2005). Wittchen and colleagues (2002) found that recognition rates by primary care practitioners were only 34.4% for pure GAD and 43% for GAD with comorbid depression. There are likely to be a variety of reasons why GPs are poor at recognising anxiety disorders in their patients. People with GAD may have symptoms of anxiety, worry, tension, irritability or tiredness, about which they feel reluctant to complain to their GP because they do not view these symptoms as being ‘medical’, or the GP may identify these as symptoms of a more general malaise and not specifically consider or ask about anxiety as a possible cause (Arroll & Kendrick, 2009). In addition, many people may present with somatic symptoms associated with their anxiety, considering these to be more legitimate or more troubling. It appears that people with anxiety disorders are often frequent users of primary care resources, but if the anxiety component of their problem is not detected they may not receive the correct treatment and may undergo unnecessary and costly investigations, in particular for their physical symptoms (Hales et al., 1997). Recognition is increased by factors such as older age, presentation of other psychological problems, and enhanced knowledge, skills and attitudes of practitioners in primary care (Tylee & Walters, 2007).

There is evidence that GPs may not offer effective evidence-based treatments to people with anxiety disorders as often as may be indicated, and that the treatments offered are more likely to be pharmacological, rather than psychological therapies such as cognitive behavioural therapy (CBT) (Stein et al., 2004) due to limited availability of such treatments, although this may be changing with increased access to psychological therapies through the Improving Access to Psychological Therapies programme (IAPT).2 The majority of treatments offered for anxiety disorders are likely to be based in primary care and may involve the GP and/or a low-intensity psychological therapist such as a primary care mental health worker or the practice counsellor. Self-help bibliotherapy and web-based interventions may be effective for some people with GAD, although referral to secondary care practitioners, such as a high-intensity psychological therapist, may occur for those more severely affected. Referral to secondary care psychiatric mental health services is likely to be rare and reserved for people with the most treatment-resistant symptoms and severe functional impairment.

In summary, there is evidence that GAD is currently significantly under-detected and under-treated in UK primary care settings. This is a potentially serious omission, given the functional impairment and chronicity that can be associated with this diagnosis, particularly when comorbid with depression or physical health problems. There needs to be an increased emphasis on encouraging people to actively present their anxiety symptoms, and for their GPs to be more attuned to this diagnosis (particularly in people known to have depression or a chronic physical health problem) and the need to provide effective evidence-based treatments as early as possible in the course of this disorder before it becomes a long-term problem.

2.4.2. Assessment and co-ordination of care

Primary care and mental health practitioners need to have skills in the identification of GAD and its differentiation from other anxiety and depressive disorders in order to assess GAD and provide appropriate treatment. Assessment involves evaluation of GAD symptoms, especially worry and somatic symptoms of anxiety, the duration of these symptoms, and the extent of the person’s functional impairment and distress and their coping resources. Assessment also needs to include evaluation of the symptoms of other anxiety and depressive disorders (especially panic disorder, hypochondriasis, OCD, social phobia, major depressive disorder and dysthymic disorder) given both the overlap of symptoms (for differential diagnosis) and the comorbidity between GAD and these other disorders.

The majority of treatment takes place in primary care or is linked with primary care, usually by either being directly provided by GPs or by psychological practitioners in liaison with GPs. GPs are accordingly central to the coordination of care. Ensuring a clear collaborative treatment plan between GP and psychological practitioners is important. For a small minority of people with very severe disorders, treatment may be provided by a multi-professional team in secondary care with coordination of care through the Care Programme Approach (CPA).

2.4.3. Aims and non-specific effects of treatment and placebo

The aim of treatment for GAD is to relieve symptoms, restore function and prevent relapse. The latter goal is important because GAD manifests as a chronic, relapsing condition and recurrence of illness is common, even when short-term treatment has apparently been successful (Yonkers et al., 1996). In clinical trials, the outcome of treatment is often determined on standardised rating scales and can be divided into ‘response’ (where the symptom score has dropped by at least 50%) and ‘remission’ (almost complete relief of symptoms). In the treatment of depression, remission rather than response is now seen as the preferred goal because people who are essentially asymptomatic have improved functional outcomes and less risk of relapse. It seems probable that similar considerations will apply to the treatment of GAD.

Many people with GAD have had symptoms for long periods of time. Nevertheless, in short-term studies of medication, pill placebo treatment in the context of the clinical care provided by a controlled trial is certainly beneficial for a proportion of people. For example, in a 12-week placebo-controlled trial of escitalo-pram and paroxetine, just over 40% of participants responded to placebo and about 30% reached remission (Baldwin et al., 2006). In contrast, naturalistic follow-up studies of people with GAD in the community have found considerably lower remission rates than this, at about 15% a year (Yonkers et al., 1996). This suggests that either GAD, despite its chronicity, can respond well to pill placebo and non-specific aspects of good clinical management, or that the people who participate in placebo-controlled trials are not typical of the broad range of people with GAD in the community. In addition, it is not known whether people who respond to a placebo in the short-term will maintain this level of improvement whereas there is some evidence that continuing drug treatment that proved effective in the short-term can help prevent relapse (Baldwin et al., 2005).

Non-specific effects of treatment are also important in assessing the benefits of psychological therapies such as CBT and applied relaxation. Often such treatments are assessed against ‘waitlist’ or ‘treatment as usual’ control groups, which means that the non-specific effects of factors such as increased professional support and instillation of hope will augment the specific effects of a particular therapy. Thus a meta-analysis showed that while CBT was superior to waitlist control in the treatment of GAD, its superiority to supportive psychological therapy could not be clearly demonstrated (Hunot et al., 2007).

Consistent with this, a substantial number of other approaches have been employed to help people with anxiety disorders, such as exercise, prayer and homeopathic and herbal remedies (Jorm et al., 2004). This suggests that numerous non-medical approaches, provided they carry meaning and hope for the person concerned, can enable individuals to use their own coping and healing capacities to overcome anxiety symptoms. At present it is not possible to identify those people who will respond to non-specific, as opposed to specific, pharmacological and psychological treatments. In the treatment of depression it appears that the response to placebo lessens as the condition becomes symptomatically more severe (Khan et al., 2005); this means that the specific benefits of antidepressants are greater in the most severely ill people. Whether the same is true in people with GAD is not clear.

2.4.4. Pharmacological treatments

Placebo-controlled trials indicate that a wide range of medicines with differing pharmacological properties can be effective in the treatment of GAD (Baldwin et al., 2005). Traditionally, benzodiazepine drugs, such as diazepam, were employed for this purpose but it became clear that their use was commonly associated with the development of tolerance and dependence (Royal College of Psychiatrists, 2005). For this reason they are now recommended only for short-term use (2 to 4 weeks). Another drug specifically licensed for the treatment of GAD is buspirone, which acts on a particular subtype of serotonin receptor. However, like benzodiazepines, buspirone is recommended for short-term use only (British Medical Association & the Royal Pharmaceutical Society of Great Britain, 2009).

In recent years antidepressants such as SSRIs have been increasingly used to treat GAD (Baldwin et al., 2005). Unlike benzodiazepines, antidepressants do not relieve anxiety from the beginning of treatment and a period of some weeks often needs to elapse before significant clinical improvement is seen. Tolerance and dependence do not seem to be a problem with antidepressant treatment, though it should be noted that, like benzodiazepines, antidepressants can cause discontinuation symptoms on abrupt withdrawal (MHRA, 2004). As well as SSRIs, serotonin noradrenaline reuptake inhibitors (SNRIs), such as venlafaxine and duloxetine, are also effective in GAD, as are the older and less selective tricyclic antidepressants (TCAs), such as imipramine. However, TCAs are not as well tolerated as newer antidepressant agents and are more dangerous in overdose (Baldwin et al., 2005).

In addition to the antidepressants, other compounds also have efficacy in the treatment of GAD. These include the antihistamine hydroxyzine, and the anticonvulsant drug pregabalin, which binds to a subtype of calcium channel in the brain (Baldwin et al., 2005). Both conventional antipsychotic drugs and the newer ‘atypical’ antipsychotic agents have also been used in the treatment in GAD, both as a sole therapy and as an ‘add-on’ to SSRI therapy when the latter has proved ineffective (Pies, 2009). However, the greater side-effect burden of antipsychotic drugs means that their use is currently restricted to people with refractory conditions, with prescribing guided by secondary care.

While many drug treatments have been demonstrated to be effective in GAD relative to placebo, there are very few comparative studies between active pharmacological agents. In addition there are no reliable clinical or biological predictors of treatment response in individuals. For this reason the selection of pharmacological treatment is usually made on the basis of the side-effect profile and the history of medication response in a particular individual.

2.4.5. Psychological treatments

Developments in psychological treatments for GAD have tended to parallel changes in the conceptualisation and diagnostic criteria for GAD, moving from a more general approach to more specific interventions.

Early psychological treatments for GAD tended to involve non-specific interventions such as supportive psychotherapy and relaxation training. Initial cognitive behavioural packages for the treatment of GAD (Borkovec & Costello, 1993; Barlow et al., 1992) focused on the treatment of persistent anxious arousal and often included a number of interventions such as applied relaxation, imagery rehearsal (imaginal practice of coping skills in response to anxiety), stimulus control (establishing increased control over worry) and cognitive approaches based on the work of Beck and colleagues (1985).

More recent adaptations of CBT have emphasised the specific role of worry in GAD and have tried to focus treatment more on the processes thought to underlie the disorder. An example of this is CBT targeting the intolerance of uncertainty (Dugas et al., 2007) or the metacognitive therapy developed by Wells (1999), which emphasises the importance of the beliefs people have about worry and attempts to modify these.

Borkovec and colleagues (2002) have augmented existing CBT protocols with interpersonal/psychodynamic strategies to address problematic relationship patterns often found in people with GAD and the implications of the avoidance theory of worry, suggesting that people with GAD worry in order to avoid experiencing negative emotions.

Other adaptations of CBT have integrated acceptance-based and mindfulness approaches into treatment for GAD, incorporating the acceptance and experience of frequently avoided emotions into treatment protocols (Orsillo et al., 2003).

2.4.6. Stepped care

Stepped care (Scogin et al., 2003) is a framework that is increasingly being used in the UK to specify best practice in the design of clinical pathways to care. Stepped care is designed to increase the efficiency of service provision and therefore benefit patient populations. The basic principle is that patients presenting with a common mental health disorder will ‘step through’ progressive levels of treatment as necessary, with the expectation that many of these patients will recover or improve while undergoing less intensive treatments. The key features of stepped care are that treatments delivered first should be the least restrictive and that the model is self-correcting. The definition of ‘least restrictive’ may refer to the impact on patients in terms of cost and personal inconvenience, but can also refer to the amount of specialist therapist time required (that is, treatment intensity). High-intensity treatments are reserved for patients who do not benefit from low-intensity treatments, or for those who can be accurately predicted to not benefit from such treatments. ‘Self-correcting’ in this context means that the decisions about treatment provision and the effects of treatment are monitored systematically, and changes are made (‘stepping up’) if current treatments are not achieving significant health gain. Thus, stepped care has the potential for deriving the greatest benefit from available therapeutic resources (Bower & Gilbody, 2005).

Successful implementation of a stepped-care model is crucial for effective implementation of the NICE guidelines (Lovell & Bee, 2008). There are two conceptualisations of the stepped-care model. The first is a sequential model, where all people move through the steps in a systematic way, regardless of severity, need or choice. All patients initially receive an evidence-based low-intensity treatment and only ‘step up’ if and when they have not benefited from the low-intensity treatments offered. The second model is a stratified or multiple-access model, which allows patients to access more intensive treatment initially, without having received less intensive interventions first (Lovell & Richards, 2000). Stratified stepped-care models have been incorporated into previous NICE guidelines, where stratification has been determined by the person’s degree of functional impairment (as in the NICE guideline on OCD and body dysmorphic disorder; NICE, 2005b) or severity of the disorder (as in the NICE guidelines on depression; NICE, 2009b; 2009c).

2.4.7. The economic cost of anxiety disorders – focus on generalised anxiety disorder

Anxiety disorders place a significant burden on individuals as well as on the health-care system. Andlin-Sobocki and colleagues (2005) estimated the cost of anxiety disorders in Europe using published epidemiological and economic data from 28 European countries. Data on healthcare resource utilisation (medication, hospitalisation and outpatient care) and productivity losses due to sick leave associated with anxiety disorders were based on a German national health survey. The estimated total cost of anxiety disorders in Europe was reported to reach €41 billion (2004 prices). The average annual additional cost per person with GAD (relative to a person without an anxiety disorder) was estimated at €1,628 in 2004; of this, 76% was associated with provision of healthcare services and the remaining 24% with productivity losses due to sick leave (Andlin-Sobocki & Wittchen, 2005). The additional per-person cost of GAD was found to be the highest among respective costs of other anxiety disorders, such as panic disorder, agoraphobia, social phobia and OCD.

Only limited data on the healthcare resource utilisation by people with anxiety disorders exist in the UK. According to the Hospital Episode Statistics, in the financial year 2007 to 2008, 8,682 admissions were reported for phobic and other anxiety disorders in England, resulting in 121,359 inpatient bed days; of these, 747 admissions and 16,733 bed days were attributed specifically to GAD (NHS, The Information Centre, 2009). According to the most recent Adult Psychiatric Morbidity in England survey (McManus et al., 2009), only 34% of people with GAD were receiving any kind of treatment for their condition at the time of the survey. Of them, 53% were receiving medication, 21% counselling or other psychological therapy, and 26% a combination of drugs and psychological treatment. In addition, 1% of respondents with GAD reported that they had used inpatient services for their condition over the past 3 months, 8% had used outpatient services during the same period, while 25% had used community or day care services during the past year.

A number of studies have estimated the cost of anxiety disorders in the US. DuPont and colleagues (1998) estimated this cost at $46.6 billion in 1990, which accounted for 31.5% of the total cost of mental disorders in the country. The estimated cost was incurred by healthcare resource utilisation such as mental health services, medication, hospitalisation, nursing homes and outpatient visits (23.1%), productivity losses (76.1%) and, to a lesser extent, by provision of other services such as criminal justice services, incarceration, social welfare administration, as well as family care-giving (0.8%). Greenberg and colleagues (1999) provided a more up-to-date figure of the cost of anxiety disorders in the US, at $63.1 billion in 1998.

A retrospective, multivariate analysis of data derived from a large claims database in the US demonstrated that people with anxiety disorders are more likely to use outpatient mental health services compared with a control group; they are also more likely to visit medical specialists such as cardiologists and neurologists and to use hospital services, including accident and emergency services. Furthermore, compared with controls, people with anxiety disorders were found to miss more days of work or to have a short-term disability (Marciniak et al., 2004). According to the same analysis, the total medical cost per person with any anxiety disorder was estimated at $6,475 in 1999 (Marciniak et al., 2005). The multivariate model indicated that, controlling for demographics and other disease states, GAD was associated with an increase of $2,138 in the total medical cost per person.

An Australian study (Andrews et al., 2004) estimated the total annual cost of routine treatment for GAD in Australia at AUS$112.3 million in 1997 prices, based on the results of a national survey of mental health and wellbeing, and an estimated treatment coverage of only 38%. By applying optimal treatment (as achieved by operationalising detailed clinical practice guidelines and expert reviews) and increasing treatment coverage to 70%, the total annual direct medical cost of GAD was expected to rise to AUS$205.1 million.

Anxiety disorders are associated with a wide range of comorbidities, which result in a substantial increase in total healthcare costs. Souêtre and colleagues (1994) estimated the total direct and indirect costs incurred by people with GAD, with and without comorbidities, using data on 999 people participating in a French cross-sectional study. Controlling for confounding variables, the prevalence of healthcare utilisation in terms of hospitalisation, laboratory tests and medications and the respective medical costs were found to be significantly higher in people with GAD and other comorbidities, as opposed to those with GAD without comorbidities. Moreover, comorbidities were associated with increased absenteeism from work. In particular, comorbid depression (Marciniak et al., 2005; Wetherell et al., 2007; Zhu et al., 2009) and physical pain (Olfson & Gameroff, 2007; Zhu et al., 2009) have been found to have a significant impact on treatment costs incurred by people with GAD.

Efficient use of available healthcare resources will maximise the health benefits for people with GAD and can potentially reduce costs to the healthcare system and society in the long term.

Copyright © 2011, The British Psychological Society & The Royal College of Psychiatrists.

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Enquiries in this regard should be directed to the British Psychological Society.

Cover of Generalised Anxiety Disorder in Adults
Generalised Anxiety Disorder in Adults: Management in Primary, Secondary and Community Care.
NICE Clinical Guidelines, No. 113.
National Collaborating Centre for Mental Health (UK).
Leicester (UK): British Psychological Society; 2011.

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...