Anxiety in Later Life

Anxiety disorders occur in 8% of older adults (1) and, although more common than mood disorders, continue to receive less attention. Anxiety disorders are associated with low socioeconomic status, female gender, comorbid general medical illness, and disability (2). Overlapping symptoms of general medical illness and anxiety disorders confound diagnosis, resulting in underdiagnosis. Furthermore, older adults are less likely to accurately identify anxiety than are younger adults (3), and anxiety disorders are often comorbid with depressive disorders (4). Use of measures that are well validated to detect clinically significant anxiety of older adults (5), such as the Beck Anxiety Inventory (6) and the Geriatric Anxiety Inventory (7), improve recognition and treatment.

Any new anxiety symptoms in later life should prompt an investigation for medical causes. Cardiopulmonary illnesses that are associated with anxiety include congestive heart failure and arrhythmias. Endocrine illnesses include diabetes and thyroid disease. Medications, including asthma medications (e.g., albuterol) and steroids, often precipitate anxiety. Caffeine—which is in some medications (e.g., Anacin) and in tea, coffee, and soft drinks—can also cause anxiety in higher doses.

This article aims to improve the recognition and management of common anxiety disorders and presentations among older adults through review of the disorders’ epidemiology, phenomenology, and treatment. Posttraumatic stress disorder and obsessive-compulsive disorder will not be discussed in this review, as neither disorder is classified under anxiety disorders in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.).

Generalized Anxiety Disorder

After phobias, generalized anxiety disorders (GADs) are the most common anxiety disorder of older adults, occurring in 1.2% to 4.6% of those in a community-based sample. Among those with the disorder, onset occurs in late life (after age 55) for about 25% (1). The objects of excessive worry, the hallmark of GAD, most often include health, disability of self or spouse, and finances (8).

Although findings from numerous trials of pharmacologic agents are heterogeneous, the evidence of their effectiveness is established (9), with recommendations favoring selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) for older adults (10). Benzodiazepines should be avoided because of increased risk of falls and cognitive impairment in older age groups. A meta-analysis including 41 studies with 2,132 patients found strong evidence for the effectiveness of psychotherapy for GAD (11). Cognitive-behavioral therapy (CBT) was the most common psychotherapy and had large treatment effects in studies that included a control group for comparison.

Panic Disorder

Panic disorder (PD) occurs in less than 0.5% of older adults, and onset in late life is rare (12). Prevalence ranges vary, from 0.4% to 2.8%, depending on the instrument used across studies (13). The majority of cases represent chronic PD. Thus, new-onset PD in the elderly population should prompt a search for underlying medical conditions or culprit medications, including withdrawal syndromes. As with younger patients, comorbid depressive and other anxiety disorders are common. Symptoms are less severe for older than for younger individuals (14). A small, randomized controlled trial comparing paroxetine, CBT, and wait-listing for patients older than 60 years showed superior response to CBT for those with late-onset and shorter illness duration PD (15). SSRIs remain the recommended first-line pharmacological agents for older adults with PD, primarily because of their side-effect profile (12).

Social Anxiety Disorder (Social Phobia)

In comparison with younger adults (ages 17–55 years), older people have overall fewer symptoms of social phobia (16). On the Social Phobia and Anxiety Index, older adults scored higher in only two categories: “anxious when talking about business” and “anxious when writing or typing in front of others.” In their binary logistic regression analysis, Miloyan et al. identified a core set of six distinct symptoms associated with social phobia across the adult lifespan, irrespective of age. These symptoms are “thinking about social situations almost always made you anxious”; “usually became upset/anxious when you had to be in social situations”; “remained in social situations because you had to be there, even though it made you anxious”; “avoid social situations because of strong fear of them”; “thought you were more anxious about social situations than most people”; and “thought own fear/avoidance of these social situations was stronger than it should be.”

Given the paucity of data for treatment of social anxiety disorder of older adults, treatment recommendations are based on those for the adult population. In a recent meta-analysis by Mayo-Wilson et al. (17), individual CBT and SSRI and SNRI medications had greater effects on treatment outcomes than did appropriate placebos. In the absence of a single medication’s being superior in treatment, medications should be selected on the basis of side-effect profiles and tolerability. Combinations of psychotherapy and pharmacology were not found to be superior to psychotherapy or pharmacological interventions alone. However, d-cycloserine used as an adjunct with psychotherapy was shown to have an additional benefit in comparison with psychotherapy alone (18). A recent meta-analysis found that d-cycloserine enhances exposure therapy and is effective when administered at a time close to the exposure therapy, at low doses and a limited number of times. The meta-analysis included two studies on obsessive-compulsive disorder, two studies on PD, and two studies on social anxiety disorder (19). Psychological interventions were recommended as initial treatments because the effects of psychological treatments were found to be generally well maintained at follow-up (more so than medications) and because there was a lower risk of side effects with psychotherapy.

Specific Phobia

The prevalence of specific phobias in late life varies in the literature. Two studies (20, 21) have found the prevalence of specific phobias among elderly adults to be 11.5%. In a longitudinal study of 558 persons in the general population (22), the prevalence of specific phobias decreased from 9.9% at age 70 years to 6.9% at age 75 years, and eventually to 4.0% at age 79 years. Sigström et al. (22) suggested several predictors of persistent specific phobias, including female gender, comorbid anxiety disorder or depressive disorder, and certain fears (animals, natural environment, specific situations, and blood-injection injury).

Kogan and Edelstein (23) found that many specific phobias are age specific and not addressed in available screening instruments. Accordingly, they developed the Fear Survey Schedule—II for Older Adults (FSS-II-OA) by modifying the Fear Survey Schedule—II (FSS-II). Of the 22 listed fears, 11 are age specific. The fears added in the OA version are mental decline, poor well-being of loved ones, inability to care for oneself, diminished health, being robbed or attacked, falling, being physically disabled, being a burden, losing sight, losing hearing, and incontinence. On the basis of Kogan and Edelstein’s studies, the FSS-II-OA was found to have high levels of internal consistency and test-retest reliability for total scores.

People with phobias may be misdiagnosed as having PD because those with specific phobias frequently have symptoms commonly experienced in panic attacks (24). Two features distinguish PD from specific phobias, that is, people with specific phobias do not have trait anxiety (25) and, whereas those with PD are unsure about the specific moment during which they will experience a panic attack, people with specific phobias have an external, circumscribed stimulus that triggers anxiety (24).

There is a dearth of data regarding treatment of specific phobias in elderly samples. However, when we drew from the adult literature, we found that exposure therapy has had the greatest efficacy for the variety of phobias that people may experience. d-cycloserine has shown promise as an adjunct for psychotherapy. Benzodiazepines have been shown to be helpful only in relieving acute anxiety but not in cases of repeat exposure to the phobic stimulus (26). Two small, randomized controlled trials, by Alamy et al. (27) and Benjamin et al. (28), found escitalopram (5–20 mg/day) and paroxetine (20 mg/day) effective for treating the symptoms of specific phobias. However, these authors noted that larger studies are needed to confirm the studies’ findings.

Substance- and Medication-Induced Anxiety

Several substances have known anxiogenic properties, with substances—including yohimbine and carbon dioxide—even being used in research specifically because of the predictable way in which they can induce anxiety in vulnerable populations. In addition to medications, many known substances of abuse induce anxiety (Box 1), either during intoxication or withdrawal (Box 2) (29). The mechanisms by which various substances are anxiogenic are diverse and include reduced functioning of gamma-aminobutyric acid receptors, increased production of norepinephrine or other neurotransmitters, hormonal changes including changes in the hypothalamic-pituitary-adrenal axis, and so on (30). A thorough review of the medication list, including over-the-counter and herbal medications, can be a vital and revealing part of the evaluation of patients presenting with anxiety.

Fear of Falling

Fear of falling is an intense fear of standing or walking, with an ongoing concern about falling that leads to an individual’s avoiding activities that she or he remains capable of performing (31). It is one of the most common anxiety conditions of older adults. In a review of studies (32), the prevalence of fear of falling ranged between 20.8% and 85%. In the studies that included older adults who had no history of a fall, of those surveyed, 50% with fear of falling had not experienced a fall. As people age, their risk of falling increases (33) and with it, worries about falling. However, among older adults who have a fear of falling, the anxiety is excessive and debilitating. Most studies reported at least one fall as being an independent risk factor for developing fear of falling. Other risk factors include increasing age, female gender, dizziness, self-rated health status, and depressive disorder. Finally, fear of falling has been shown to have a negative impact on physical, functional, psychological, and social domains.

Treatments for fear of falling include exercise interventions, hip protectors, and combined efforts. Meta-analyses suggest low significant effects for these interventions, though tai chi has shown moderately significant effects (34). Levy et al. (31) took a CBT approach to the treatment of fear of falling and developed a virtual-reality intervention. Through progressively more challenging levels, participants learned to overcome their fear of falling. The authors concluded that the “virtual reality therapy is an effective treatment for fear of falling syndrome” (p. 880).

Anxiety in Neurocognitive Disorders

Anxiety is a well-known feature associated with multiple types of neurocognitive disorders, with studies focused on the behavioral and psychological symptoms of dementia representing a relatively large body of literature. Prevalence estimates of anxiety in dementia range from 17% to 52% (35). Less is known about anxiety presenting as a premorbid or early sign of a neurodegenerative process. Because 99% of anxiety disorders start prior to the age of 65 years (36), late-onset anxiety should prompt a careful search for cognitive decline or other etiologies. Patients presenting with anxiety symptoms and mild cognitive impairment were almost twice as likely to convert to a diagnosis of Alzheimer’s disease on three-year follow-up as patients having mild cognitive impairment without anxiety (37). However, the relationship between anxiety and neurodegenerative disease is complicated, and causality is difficult to determine, given that some, although limited, evidence suggests that patients with life-long anxiety or depressive disorders are more likely to develop a neurocognitive disorder (38–40).

The Rating Anxiety in Dementia scale and the Neuropsychiatric Inventory–Anxiety subscale are tools used to measure anxiety in patients with cognitive impairment. Self-rated anxiety also can be measured with scales such as the Geriatric Anxiety Inventory. A Cochrane review of the literature found that psychological treatments of anxiety, including CBT, are effective at reducing clinician-rated anxiety but not necessarily self-rated anxiety among patients with dementia (41).

Anxiety With Co-Occurring General Medical Illnesses

Among medical illnesses experienced by elder populations, those with chronic obstructive pulmonary disease (COPD) are more often afflicted with anxiety (42). For people with cardiovascular disease (CVD), anxiety has been shown to confer greater heart disease risk than depressive disorders (43). This section highlights the current understanding of anxiety in COPD and CVD.

In a 2013 review by Willgoss and Yohannes (44), the prevalence of anxiety in patients with COPD ranged from 13% to 46%. In addition, patients with COPD and comorbid anxiety disorders are twice as likely as those without anxiety to exhibit self-reported functional limitations, poorer exercise tolerance, and higher frequency of acute COPD exacerbations. No pharmacological or psychotherapeutic interventions have been shown to definitively improve anxiety among people with COPD (45).

As is noted above, individuals with anxiety are at greater risk for CVD. Anxiety disorders also increase the risk for subsequent major adverse coronary events, for example, myocardial infarction and stroke for persons with established CVD (46). Scherrer et al. (47) noted that identifying anxiety among people with acute coronary syndrome is especially important, because people with anxiety are more likely to have insufficiently treated depressive illnesses that further increase their risk of myocardial infarction. Treatment of anxiety with SSRIs for patients with CVD is supported by practice guidelines (48). However, SSRIs should be used with caution and must incorporate knowledge about specific agent’s (or agents’) impact on QTc. In terms of psychotherapeutic interventions, Tully et al. (49) developed a modified CBT approach, termed “panic attack treatment in comorbid heart diseases,” or PATCHD. It showed reduction in CVD hospital admissions and length of stay, panic attacks, general anxiety, and depression.

Conclusions

Most anxiety disorders of older adults are chronic conditions with onset at a younger age. However, in the case of generalized anxiety disorder, approximately half of cases have onset after the age of 50. Scales specific to anxiety of older adults may be more sensitive and accurate diagnostic tools than are measures developed for younger cohorts. As for younger adults, treatment for older adults includes SSRIs, SNRIs, and psychotherapy, including CBT. In comparison with younger and middle-aged adults, older adults with anxiety are less likely to seek treatment. More research is needed to identify the most effective treatment regimens in this age group, because current guidelines are often extracted from research on younger adults.

In the case of PD and social phobia, symptoms of older individuals tend to be less severe than they are for younger adults, although underreporting of symptoms is also present in this population.

Most anxiety disorders decrease in prevalence with age, with the exception of a few unique anxiety presentations that are much more common in the elderly population. Fear of falling has features similar to a specific phobia, and prevalence of this syndrome increases with age. Although history of at least one fall is a risk factor for fear of falling, it can occur among individuals with no history of falls. Regardless, it can have a significant impact on quality of life and independence. In addition, many patients with dementia have symptoms of anxiety. The relationship between anxiety and neurocognitive disorder is complicated and may even be bidirectional, although there is more evidence supporting anxiety as a premorbid feature of dementia rather than as a cause.

Finally, with high rates of comorbid general medical illnesses, diagnosis becomes more challenging and anxiety disorders easier to miss or misdiagnose.