Anxiety disorders are considered typical of childhood and early adulthood, but are in fact also commonly experienced in community-dwelling older adults. Approximately 15-50% experience significant anxiety, with 1.2-15% attracting a formal diagnosis of anxiety-related disorder.1 This proportion rises to over 20% in carers of people with dementia.²

Anxiety disorder in the elderly is therefore twice as common as dementia and four to six times more common than major depression.³ Anxiety is associated with poorer quality of life,⁴ significant distress⁵ and contributes to the onset of disability.⁶ Mortality risks are also increased, through physical causes, especially cardiovascular disease,^{7, 8} and suicide.⁹ The cost of late life anxiety disorders increases exponentially with age, through direct care for anxiety itself¹⁰ and as a result of exacerbation of existing medical conditions and increased use of healthcare services.¹¹ Despite their significant prevalence and impact, anxiety disorders remain underdiagnosed and undertreated.¹²

Ninety per cent of diagnosable anxiety disorders in older adults are generalised anxiety disorder (GAD) or specific phobias,¹³ with GAD accounting for more than half of all anxiety diagnoses.¹⁴ Obsessive compulsive disorder (OCD), panic disorder and post-traumatic stress disorder (PTSD) make up the other 10%.

First-onset presentations are rare at this age, with 50-97% being exacerbations of earlier onset anxiety disorders.¹⁴

Diagnosing anxiety disorders in older adults remains a challenge because of the significant overlap in symptoms between physical disorders (shortness of breath; abdominal and chest pain; palpitations) and depression (disturbed sleep; poor attention, concentration and memory; restlessness).

Presentation

GAD is the most common disorder and is characterised by excessive and difficult to control worry about a number of life domains (e.g. health, relationships, finance) of more than six months’ duration. It is accompanied by physical symptoms, such as fatigue, sleep disturbance and agitation, with physical presentations being more common in the elderly.¹⁵ These somatic symptoms frequently create diagnostic ambiguity especially when chronic physical disorders are also present.¹⁶ Depression is a major comorbity to the extent that some have questioned whether GAD is an independent syndrome in older adults.1 Importantly comorbidity is linked to greater risks for depression becoming chronic.¹⁷

Specific phobia refers to a persistent irrational fear of something that poses little or no actual danger. It is associated with loss of function through avoidance behaviour. Fear of crowds or going out is the leading focus of specific phobia in older adults (80% of new onset cases¹⁸). It is not usually a component of panic disorder as is frequently the case in younger adults. Instead it tends to follow traumatic events such as falls, physical illness or aggression from others.¹⁹

Fear of falls is an-age specific anxiety occurring in 30-77% of those who have had a fall.²⁰ It has an independent functional effect leading to restriction of social and physical activity beyond the physical effects of falls.²¹ This avoidance in turn leads to poorer quality of life,²² and increases the risk of further falls through poor balance and lack of physical conditioning.²³ Older adults also have more frequent phobias of natural phenomena (e.g. lightning, heights).²⁴

OCD, panic disorder and PTSD tend to occur in the context of physical illness or dementia. New onset idiopathic OCD is relatively rare, with the vast majority of patients graduating from general adult settings or with age-related changes limiting the individual’s ability to resist long-term subclinical obsessionality.²⁵

Preoccupations tend to shift towards sin and religion thus replacing contamination fears, symmetry concerns or counting rituals, which are more common in younger people.²⁶ New onset cases do occur prodromal to dementia, where preoccupation with newly difficult routines (e.g. toileting and medication schedules) may be the presenting feature.²⁶ Panic disorder in older adults tends to present with more shortness of breath than with other physical features.²⁷

PTSD may either occur in response to a new trauma (e.g. falls, physical illness, traumatic events) or may be a new manifestation of dormant trauma (e.g. in war veterans), which is re-activated by personal losses (e.g. bereavement, retirement), or news of war.²⁸

Examination and investigation

Good history taking is crucial in elucidating whether the complaint is of new onset or whether it is a recurrence of a previous disorder.

The presence of comorbid depression should be clarified. If present, its temporal relationship with the anxiety symptoms will indicate whether there is an independent anxiety disorder.

A medication review is warranted, as a number of drugs may be causative (calcium channel blockers, alpha- and beta-blockers, digoxin, L-thyroxine, bronchodilators, steroids, theophylline, antihistamines) or may cause anxiety in withdrawal (e.g. benzodiazepines).²⁹ Substance and alcohol abuse should be excluded, as withdrawal from either may cause anxiety.

A new or exacerbated physical illness may be related to anxiety, and patients should undergo basic investigations to exclude: ²⁹

• Anaemia

• Cardiovascular and cerebrovascular pathology

• Diabetes and hypoglycaemia

• Hyperthyroidism

• Hypocalcaemia, hyponatraemia, hyperkalaemia

• COPD

Medical investigations will help clarify the extent to which a particular somatic symptom is the result of anxiety. Longitudinal follow-up of patients and their response to adequate treatment will be key in confirming the diagnosis. 

Measuring anxiety

Measures of anxiety are often developed with the general adult population in mind. They are less valid when applied to older adults, because of reliance on somatic symptoms and questions having multiple possible answers.

The Geriatric Anxiety Inventory was developed specifically for the older adult population and is relatively brief (20 items), provides only dichotomous choices (agree or disagree) and probes a very limited number of somatic symptoms.³⁰ A score of 11 indicates a high  likelihood of GAD, identifying 83% of GAD patients correctly (sensitivity 75%, specificity 84%), while a score of 9 confers a high likelihood of any anxiety disorder (correctly classifies 78% of patients; sensitivity 73%, specificity 80%).

The four-item FEAR questionnaire has been developed specifically for assessment of elderly primary care attenders, see table 1, opposite.31 It can be used to clarify the diagnosis in a clinical consultation, exploring dimensions that may not be covered in a routine interview. It is a shortened version of the Anxiety Disorder Scale (ADS), which had good validity and inter-rater reliability in detecting anxiety disorders in a community sample aged over 65 years.³²

The FEAR questionnaire includes four items of the ADS probing:

• Frequency of anxiety

• The enduring nature of anxiety

• Use of alcohol or sedatives to reduce anxiety

• Presence of restlessness or fidgeting

A score of 2 or more predicts a high likelihood of an anxiety disorder (sensitivity 77% and specificity 83%).

Management

Most patients are effectively treated within primary care, using primary care access to psychological therapy, such as Improved Access to Psychological Therapies (IAPT).

Referral to secondary services, community mental health teams, may be appropriate if the first or second line of treatment fails, there is comorbid major depression, which worsens the treatment prospects, or if there are significant concurrent medications with the risk of interactions (e.g. inhibition of cytochrome P450 by SSRIs leading to increased plasma levels of drugs metabolised by P450, such as phenytoin, warfarin, steroids, theophylline and methadone; risk of exacerbating upper GI bleeds with SSRIs).

Psychiatric teams manage anxiety patients generally in an outpatient setting with hospital admissions confined to instances where there is acute suicidality or risk of self-neglect. 

The evidence for treatment in older adults is limited because randomised controlled trials largely exclude people over the age of 65. Recommendations are frequently derived from working age adults with mainly general applicability to older adults. The principles of pharmacotherapy in this age group are to start in doses of half or a quarter of the normal adult starting dose and titrate slowly.²⁷

Frequent follow-up for reassurance is crucial, particularly in the group of anxiety patients who tend to be more vigilant of side effects and have a low threshold for treatment discontinuation. Benzodiazepines are still relatively frequently prescribed,³³ but come with serious side effects such as worsening of cognitive function and risk of life-threatening falls.

A meta-analysis of the available studies in older adults found that CBT and pharmacotherapy had large effects on treating anxiety relative to placebo with improvement still present at 37 weeks.³⁴

The evidence-based treatments for GAD based on studies in older adults are SSRIs, with best evidence for citalopram and sertraline,^35,36 and venlafaxine.³⁷ While pregabalin has gained a place as first-line treatment in general anxiety disorder³⁸ the evidence in older patients is promising, but limited.³⁹ Evidence for age-specific options in GAD treatment resistance is lacking,⁴⁰ as is evidence for efficacy in the other disorders in older people. Therefore first-line treatments are based on experience from working age adults (SSRIs and SNRIs). Clomipramine and other tricyclic antidepressants have traditionally had good effects in the treatment of OCD but they should be used with caution in older adults because of cardiac and severe anticholinergic side effects.   

Cognitive behaviour therapy (CBT) is an important treatment option, but there is no evidence that combining pharmacotherapy with CBT leads to an augmented response.³⁸

CBT and also relaxation training and supportive therapy have been shown to have independent positive effects on anxiety in older adults, as well as on limiting the use of benzodiazepines.^5,41,42 CBT models that account for the difficulties in comprehension and retaining information in older adults have been developed (e.g. simplify treatment rationale and interventions, between sessions telephone reminders) and a pilot trial reported improved anxiety scores with the modified therapy.⁴³

Conclusion

Anxiety disorders are common in older age and carry significant risks for deterioration in quality of life and overall physical health. Diagnosis has traditionally been hampered by the large overlap between anxiety symptoms and those associated with depression and physical illness. Primary care plays a crucial role in identifying patients and initiating timely treatment, an intervention with the potential for a wide-ranging impact.