Description of the condition

Anxiety and related disorders are the most prevalent mental disorders overall across the life span. Epidemiological studies (including, as the most important, panic disorder, agoraphobia, generalised anxiety disorder, social anxiety disorder, specific phobia, obsessive‐compulsive disorder, and post‐traumatic stress disorder) report a lifetime prevalence rate among the general population ranging from approximately 17% in the European Union (Wittchen 2011) to almost 29% in the USA (Kessler 2005). An age‐independent systematic review has found both an intercontinental difference in prevalence and a difference between countries (Baxter 2014). Between East Asia and North Africa, for example, a threefold difference in prevalence was found of 2% and 6% respectively. The parts of the world with the most epidemiological data, i.e. North America and Europe, showed differences between countries of more than 50% (for instance 2% in Israel versus almost 7% in France). The prevalence and burden of disease was also highest in the lower age groups. Nevertheless, despite these difference between studies (probably due to different methodologies), countries and continents, anxiety and related disorders are the most prevalent disorders at a global level and over the course of life.

The estimated prevalence rates of anxiety and related disorders among older adults living in the community varies across several epidemiological studies from 6% to 12%, probably due to methodological issues (Bryant 2008; Byers 2010). A meta‐analysis of epidemiological studies (Volkert 2013) on current and lifetime prevalence of anxiety disorders in older age in Western countries revealed the following figures: a current and lifetime prevalence of 0.9% and 2.6% respectively for panic disorder; 0.5% and 1.0% for agoraphobia (with and without panic disorder); 4.5% and 6.7% for specific phobia; 1.3% and 5.1% for social anxiety disorder; and 2.3% and 6.4% for generalised anxiety disorder. These figures correspond with the recently published MentDis_ICF65+ Study, with a current and lifetime prevalence of anxiety disorders of 11.4% and 21.4% respectively in Europe (Andreas 2017). This study also included post‐traumatic stress disorder, with a current prevalence of 1.4% and a lifetime prevalence of 2.5%. The prevalence of obsessive compulsive disorder in the elderly is highlighted in other epidemiological studies, with prevalence rates between 1.5% and 2.9% (Klenfeldt 2014; Préville 2010).

These figures show that anxiety and related disorders are also common mental disorders in the elderly, with prevalence rates comparable to the younger‐age categories. The diagnostic criteria of these disorders are specified in the Diagnostic and Statistical Manual of Mental Disorders (DSM‐5; APA 2013), or the International Classification of Diseases and Related Health Problems, 11th revision (ICD‐11; WHO 2018). Both international classification systems distinguish three separate sections: namely, 'pure' anxiety disorders, stressor‐ and trauma‐related disorders, and obsessive‐compulsive or related disorders. Neither the DSM‐5 nor the ICD‐11 distinguish between younger and older adults in the diagnostic criteria for these disorders. Although there are no major age‐related diagnostic differences, there are a number of characteristics that are specific to the elderly with anxiety and related disorders, and which it is relevant to review briefly.

When these disorders are diagnosed at an older age, they are generally characterised by a relatively young age at onset and a chronic course, with onset above 65 years of age a rare phenomenon (Kessler 2005). Older adults with anxiety and related disorders are at risk of being undiagnosed and untreated. The odds ratios of receiving specialised mental health care were 2.4 to 4.8 greater in the younger adults (Wang 2005). A major feature at the behavioural level between older and younger adults is the difference in helpseeking. The percentage of adults aged 65 and over seeking help is almost 50% lower compared with their younger counterparts (MacKenzie 2012). Negative stereotyping, also known as ageism, of healthcare professionals, older people themselves and their social environment contributes to this phenomenon of underdiagnosing and undertreating, by interpreting specific anxiety symptoms as appropriate for age (e.g. Bodner 2018; Wolitzky‐Taylor 2010).

Phenomenological differences can also play an important role in why anxiety and related disorders in the elderly are not recognised. Older adults are more likely to suffer from somatic morbidity, which may be given greater priority in diagnosis and treatment by healthcare professionals, and as a result of which comorbid anxiety disorders may be overlooked. Furthermore, elderly people are more inclined to translate anxiety symptoms into physical complaints, causing the focus to be mainly but mistakenly directed to these symptoms (Wolitzky‐Taylor 2010). The presentation of anxiety symptoms seems to be subject to the process of aging. Higher age in general appeared to be accompanied by a decrease in both cognitive and emotional anxiety symptoms, as well as physiological symptoms. Feelings of insecurity and stressful events may therefore be better tolerated in older age (Gould 2010). On the other hand, dysfunctional fear‐related cognitions also seem to change in older people with anxiety disorders compared to younger adults. For example, the level of agoraphobic cognitions was less severe in older adults suffering from panic disorder compared to younger adults, whereas both the severity and the nature of agoraphobic avoidance did not differ (Hendriks 2010). It was therefore suggested that in older adults suffering from anxiety disorders focusing on avoidance behaviour and associated limitations in daily functioning may be more appropriate (Grenier 2011; Hendriks 2010). Elderly people with generalised anxiety disorder experienced on average fewer symptoms compared to younger people, although specific symptoms like dizziness, decreased concentration and some physical symptoms (nausea, gastric problems) were more common in older compared to younger people (Miloyan 2014b). Differences were also apparent in worry content. Older people worried about health issues and the well‐being of relatives, whereas younger people worried more about work, social life and finances (Miloyan 2014a). In the case of social anxiety disorder no phenomenological differences between the various age groups were found (Miloyan 2014a). Finally, two types of anxiety syndrome are common and specifically related to older age: the first is fear of falling, which may be categorised as a specific phobia. Fear of falling is generally not present in young people and is characterised by changing motor skills during walking (walking more carefully, holding on to yourself, using aids such as a walker) and avoiding physical activities (Parry 2016). The second is fear of dementia, which is mainly characterised by worries about cognitive functioning and developing dementia, and is typically found in the younger elderly (Cutler 2015; Cutler 2017).

In general, people with anxiety disorders have an increased risk of comorbid disorders. Life‐time comorbidity in anxiety disorders is estimated at 80% (Brown 2001), with the Netherlands Study of Depression and Anxiety (NESDA) showing an overall comorbidity of both another anxiety disorder or depression, or both, of 70% (Hofmeijer‐Sevink 2012). Moreover, prospective epidemiological studies show that anxiety and related disorders tend to run a chronic course, meaning they are likely to resurface over the life course (Hendriks 2013; Keller 2006; Penninx 2011; Yonkers 2003).

Although the prevalence and the burden of disease in late‐life anxiety and related disorders is not as high as in younger adults (Baxter 2014), these disorders are not a minor health problem in older adults. Comorbidity of a second anxiety disorder or depressive disorder or both is common in older age (Hek 2011; Kessler 2005; King‐Kallimanis 2009; Lenze 2005; Van Balkom 2000). Secondly, physical diseases or physical complaints, which are more common in older age, are associated with an increased risk of suffering from anxiety in older age (El‐Gabalawy 2011). Thirdly, anxiety and related disorders in later life are associated with an increased risk of coronary heart disease (Huffman 2008) and a higher mortality rate (Brenes 2007; Van Hout 2004). In addition, negative stereotyping in both older adults themselves, in their social environment and among healthcare professionals was found to be a general mechanism in older age categories that contributed significantly to reducing both physical and mental health and increasing the burden of disease and healthcare costs (e.g. Levy 2020). Several studies have shown that these factors reduce quality of life in older adults suffering from anxiety and related disorders. Daily functioning decreases, well‐being is impaired, healthcare and additional costs increase, and quality of life is more affected by poor mental health compared to physical morbidity (Hohls 2018; Mackenzie 2010; Puvill 2016; Simning 2020)

Description of the intervention

First‐line treatments for anxiety disorders which do not improve after education and active monitoring consist of psychological therapy or antidepressant medication or both (NICE guideline 2011). Community studies show that only 5.1% of older adults suffering from anxiety disorders receive psychological therapy, and only 3.8% are prescribed an antidepressant (De Beurs 1999). An epidemiological study has shown that mental health services are still underused for elderly people with anxiety disorders in Europe, and that this has hardly changed during the past 20 years (8.9% in 1999 versus 10.6% in 2017) (Volkert 2017). Both participant and physician characteristics may contribute to this phenomenon. Many older adults resist the use of antidepressants, based on a fear of dependence, prior negative experiences, resistance to viewing their symptoms as a medical illness, and concern that antidepressants will prevent natural feelings (Givens 2006). Instead, they prefer psychotherapy (Luck‐Sikorski 2017; Mohlman 2012). This is important, as offering treatment tailored to the patient's preference has a sustained moderate effect in improving the outcome (Lindhiem 2014).

The average time between the onset of anxiety symptom and the first therapy contact has been shown to be between nine and 23 years, suggesting that general practitioners do not easily recognise anxiety disorders (Kroenke 2007; Wang 2005). The referral rate from primary care to a mental health specialist is three times lower for older adults suffering from anxiety disorders (17%) compared to those suffering from depression (55%) (Kessler 2005; Wang 2005). Furthermore, although pharmacological treatment with antidepressants or benzodiazepines is effective in treating late‐life anxiety disorders (Pinquart 2007), both drugs are associated with an increased risk of falling and fractures (Coupland 2011; Hartikainen 2007). As most older adults suffering from anxiety disorders are treated in primary care, general practitioners should be aware of increased side‐effects in later life, for example the heightened risk of falling and traffic accidents, memory impairment, dependency associated with the use of benzodiazepines (Maree 2016), and the low concentration of sodium in the blood with the use of serotonin reuptake inhibitors (Wright 2008). Also of concern are the adverse effects of drug interactions, such as gastrointestinal bleeding resulting from the concurrent use of serotonin reuptake inhibitors and non‐steroidal anti‐inflammatory drugs in this age group, known for polypharmacy use (Yuan 2006; Weinrieb 2005). As effective management of anxiety disorders in later life may improve outcomes, it is important that guidelines for therapy become available.

Because cognitive behavioural therapy is the first choice of psychological treatment in the current guidelines for the treatment of anxiety and related disorders in the 18‐65 age group, this review focuses on the outcome of CBT in the treatment of anxiety and related disorders in adults aged 65 years and older. 

Cognitive‐behavioural therapy (CBT) combines principles of both behavioural and cognitive therapy. These principles stem largely from learning and cognitive theories (Bandura 1969; Beck 1995; Yates 1970). Cognitive therapy (CT) assumes that a participant's symptoms (e.g. anxiety) arise from misperceptions and attitudes about the world and themselves. Behavioural therapy (BT) is a directive, active approach involving principles of learning to help the participant to develop new and adaptive ways of behaving. CBT is characterised as a structured, goal‐oriented, problem‐focused, time‐limited intervention (Dobson 2019).

Third‐wave CBT was developed some 15 years ago. It was originally designated as an evolution from the existing traditional CBT protocols to a more process oriented cognitive therapy (Dimidjian 2016). It differs from traditional CBT in that it does not focus on syndromes and reducing symptoms. The focus of third wave CBT's is, for example, on context, relationship, emotion, acceptance and the process of therapy, instead of primarily on symptoms and procedures to reduce these symptoms. These third wave CBTs have been used in both the 18‐65 age group and the 65+ age‐category (Kishita 2017). Examples of evidence‐based third‐wave CBTs are acceptance and commitment therapy (ACT), dialectical behavioural therapy (DBT), metacognitive therapy (MCT), mindfulness‐based cognitive therapy (MBCT) or schema‐focused therapy (SFT) (Hayes 2017; Kahl 2012). DBT is characterised by addressing maladaptive emotional regulation and improving tolerance of stress, regulation of emotions and interpersonal coping. ACT is directed at things patients are directly able to influence, such as their own behaviour, rather than trying to control experiences that cannot be directly influenced, such as emotions and thoughts (called experiential avoidance). This implies an accepting attitude towards these emotions and thoughts. MCT is a form of CBT for excessive anxiety. Central to this are dysfunctional thoughts about a patient's own worries. These views are called metacognitions, and are thought to underlie excessive anxiety and dysfunctional behaviours. MCT aims to change these dysfunctional metacognitions into adaptive cognitions. The central goal of MBCT is to be attentive to thoughts, feelings, physical sensations and the physical environment in an acceptable way, without judging whether something is right or wrong. MBCT helps in recognising automatic, sometimes unhelpful, patterns. Greater awareness of these patterns can lead to more freedom of choice, to breaking through automatic reactions and to learning to react in a more helpful way. SFT is a therapeutic approach in which elements from cognitive, behavioural therapeutic and psychodynamic models, among others, are combined with each other. The therapy aims to identify in patients old dysfunctional coping styles related to events and experiences in the past, and to change them to more adaptive ones; for a concise overview see Kahl 2012. Some of these therapies seem to be particularly suited to older adults with anxiety disorders (Wetherell 2011).

In order to deal with age‐associated cognitive impairment, some authors argue for the inclusion of learning and memory aids like homework reminders, troubleshooting calls or weekly review of all concepts and techniques (Mohlman 2003). They also advise the application of a more personalised approach to older people that takes into account, for example, personal context, history, attitudes towards ageing, and available skills (Laidlaw 2015). However, it is questionable whether this is a specific age‐appropriate approach or a more general approach which should always be a starting point prior to each treatment. Recent studies on older adults remain inconclusive on whether age‐related adaptations of psychotherapeutic techniques are necessary to remain efficacious for anxiety disorders in later life. Some authors argue that CBT is less effective in older people with anxiety disorders than in younger working‐age adults (Kishita 2017; Wetherell 2013). Other studies, on the other hand, argue that CBT is no less effective in older adults, and may even be more effective (Chaplin 2015; Hendriks 2014).

How the intervention might work

CBT identifies habitual ways in which participants distort information (e.g. automatic thoughts) and teaches participants to identify and respond to their dysfunctional thoughts and beliefs (e.g. catastrophic misinterpretations), using a variety of techniques to change thinking, mood, and behaviour (Clark 1986).

Behavioural interventions help the participant to develop new and adaptive ways of behaving and consists mainly of exposure‐techniques (Marks 1981). Exposure‐based therapy uses gradual, systematic, repeated exposure to the feared object or situation to desensitise participants with anxiety disorders to the feared stimulus. The controlled exposures are predictable and the participant is taught a variety of adaptive coping strategies.

CBT‐related approaches or the so‐called 'third‐wave' therapies are represented by a set of different therapies (e.g. mindfulness‐based cognitive therapy, acceptance and commitment therapy, compassionate mind training, extended behavioural activation, metacognitive therapy, schema therapy), all originating from the cognitive behavioural approach but compared to which more importance is given to the form, rather than the content, of participants' thoughts. By focusing on the function of cognition, third‐wave therapies aim to help participants to develop more adaptive emotional responses to situations. When mindfulness and acceptance are applied to anxiety disorders, the aim is for the individual to be able to observe symptomatic processes without overly identifying with them or without reacting to them in ways that cause further distress (Roemer 2008). In this review CT, BT, CBT and the third‐wave CBTs are all referred to as CBT.

Why it is important to do this review

CBT is generally considered the most effective psychological therapy for anxiety disorders, with large effects observed on the severity of anxiety symptoms, comorbid depression, quality of life and functioning (Asnaani 2020). Several authors argue that CBT should be considered the gold standard in psychological treatment (David 2018). The efficacy of CBT for anxiety disorders in young and middle‐aged adults is supported by a variety of systematic reviews and meta‐analyses (Bakker 1998; Furukawa 2006; Hofmann 2008: Hunot 2007; Loerinc 2015; Mitte 2005a; Mitte 2005b; Van Balkom 1995; Van Dis 2020; Westen 2001), although long‐term effects are not consistent across all anxiety disorders (Van Dis 2020). However, whether these results also apply to older adults remains unclear, since most trials exclude participants over the age of 65 years. It can be difficult to generalise findings from studies in young and middle‐aged adults to older adults, as older adults often have more chronic disorders which are presumed to be more resistant to successful treatment (Katz 2002). This review therefore aims to provide a comprehensive and up‐to‐date synthesis of the available evidence on CBT as a therapy for older adults with anxiety and related disorders.

Many of the earlier studies on psychological therapy for anxiety disorders in later life suggest that CBT may be effective, but have been limited to subclinical anxiety and lack rigorous inclusion criteria such as the presence of an anxiety disorder which meets diagnostic guidelines (e.g. Nordhus 2003). Previous reviews have generally concluded that psychological therapy is effective for anxiety and related disorders in older adults, but also state that there is a need for large, rigorous controlled trials for the different anxiety disorders (Goncalves 2012; Hendriks 2008; Pinquart 2007). However, in later life, pure anxiety disorders are rare, as most participants have a comorbid second anxiety disorder or depressive disorder. For the purposes of conducting this review, we will therefore combine generalised anxiety disorder, panic disorder with and without agoraphobia, agoraphobia and social phobia (conditions that often overlap), into a single diagnostic category which we term anxiety disorders. We consider obsessive compulsive disorder and the stress disorders (acute and post‐traumatic stress disorder) as separate diagnostic categories.