Sleep disturbances are very common in the context of psychological distress.1,2 Studies of individuals with insomnia have found rates of comorbid psychiatric disorders as high as 62%.3 A substantial amount of research has been conducted investigating insomnia in the context of depression in later life, but relatively little research has been conducted investigating sleep disturbances in the context of anxiety symptoms experienced by the elderly. This is surprising, given that complaints about anxiety and sleep disturbances are among the most common that the elderly express to their physicians.4 This article describes current knowledge concerning the link between anxiety and insomnia in older adults. It begins by discussing separately the research findings on sleep and anxiety, and then briefly reviews current knowledge concerning the relation between insomnia- and anxiety-related variables. It concludes with a section on the impact of the pharmacological treatment of anxiety on sleep-related variables in older patients.

SLEEP AND INSOMNIA IN THE ELDERLY
Older adults tend to go to sleep and wake up earlier than their younger counterparts.5,6 This phenomenon is commonly referred to as an advanced sleep phase. There is some evidence that, in spite of their increased likelihood of going to bed relatively early, older individuals have greater difficulty initiating sleep than their younger counterparts.7 In addition, older individuals awaken more frequently during the night than do younger persons8 and take more naps than younger adults,9 especially in the evening.10

The alternation of sleep and wake is one of the many functions governed by the suprachiasmatic nucleus of the anterior hypothalamus. There are age-related changes, such as volume and cell loss, in this structure that likely contribute to changes in sleep observed in the elderly.11 It is also likely, however, that schedule changes that accompany retirement (ie, a less structured day, fewer demands requiring wakefulness) contribute to these commonly observed changes. In addition, medical problems that can interfere with a restful night’s sleep are common in the elderly. These include but are not limited to: obstructive sleep apnea, arthritic or neuropathic pain, nocturia, and cramping related to peripheral vascular disease.

ANXIETY-RELATED DISTRESS
Anxiety disorder diagnoses are less common among older adults than among their younger counterparts.12 A number of explanations have been offered for this observation. Researchers have pointed out that the diagnostic criteria for anxiety disorders were established on the basis of symptom reports by younger adults and might not be appropriate for the elderly,13,14 and that anxiety problems that present subclinically according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV)15 criteria might have a substantial impact on functioning in the elderly.16 It has also been suggested that existing measures of anxiety might omit items that are important to the elderly, and might therefore be invalid when applied to this population.17

In spite of the evidence for a decreased prevalence of anxiety disorders in the elderly, compared to younger adults, findings indicate that anxiety-related distress may be more common in the elderly than are other forms of pathology that have received substantially more attention, including both depression and dementia.12 Himmelfarb and Murrell,18 in a study of 2860 individuals age 55 and older, found that 20% of participants reported a clinically significant level of anxious symptomatology. Female gender, lower income and education, and poorer quality housing were associated with elevated levels of anxiety. A survey of elderly persons in the Netherlands found the 6-month prevalence of anxiety disorders to be just over 10%, with women affected twice as frequently as men.19 Generalized anxiety disorder (GAD) was the most common disorder, and phobias were the next most common. Chronic medical illnesses, the prevalence of which increases among the elderly, can also contribute to anxiety in this population. For example, patients with chronic obstructive pulmonary disease may experience symptoms of anxiety and panic.20,21 Individuals in hyperthyroid states might also report symptoms of anxiety.22 Chronic heart failure has been associated with anxiety and depression.23 Similarly, patients with congestive heart failure might experience anxiety along with orthopnea. Medications can also contribute to anxiety in older adults, and the adverse impact of medications can vary across individuals. Clinicians should be watchful of some of the commonly used agents, including stimulants (eg, caffeine, theophylline), anticholinergics, antihypertensives, and digitalis. Additionally, many cold medications, including over-the-counter and “herbal” preparations, contain caffeine and/or other stimulants (eg, pseudoephedrine). Patients may also report that alcohol helps relieve their anxiety; in fact, as physiological dependence develops, alcohol can exarcerbate anxiety, depression, and even dementia, while interfering with sleep and overall wellness.

A small group of literature has developed concerning anxiety symptoms in elderly persons with dementia, but prevalence has not been clearly identified. Ballard and colleagues24 found that almost 30% of memory clinic patients with dementia had at least one anxiety symptom. Other researchers, in an observational study, determined that as many as 71% of outpatients with Alzheimer’s disease exhibit an “anxious or worried appearance,” and up to 57% demonstrate fidgeting, a form of psychomotor agitation that might be an expression of anxiety.25 Mintzer and Brawman-Mintzer26 have suggested that the assorted behavioral disturbances in this population labeled “agitation” are actually a manifestation of GAD. Other authors have critiqued clinicians and researchers for neglecting to address this important topic in greater depth.27

THE SLEEP-ANXIETY LINK IN THE ELDERLY
The literature is replete with studies investigating insomnia in the context of depressive disorders, whereas studies concerning insomnia and anxiety symptoms are less common. Nonetheless, there is empirical evidence of a higher prevalence of anxiety-related distress among individuals with insomnia. For example, Mellinger et al28 found that individuals with serious insomnia had elevated levels of a generalized anxiety–like syndrome, including somatic anxiety symptoms (eg, nervous stomach, sweating, dizziness), compared to participants with less severe insomnia. Other researchers have found that individuals with insomnia are more likely to be anxious, tense, or worried around bedtime than controls.29

A handful of studies have investigated the relation between anxiety-related variables and insomnia in the elderly. For example, Morin and Gramling30 found that older adults with insomnia complaints have higher levels of trait (ie, dispositional) and state (ie, situational) anxiety than their counterparts without such complaints. Similarly, Lichstein and colleagues31 found that, among older adults with anxiety disorders, anxiety symptoms were robustly associated with decreased sleep efficiency (ratio of time spent asleep to time spent in bed) and decreased time spent sleeping, and with increased sleep-onset latency (SOL) and time spent awake after sleep onset (WASO). Edinger et al32 found that older adults with subjective reports of insomnia demonstrate elevated levels of anxiety and dysfunctional cognitions related to sleep compared to older persons who sleep well.

Friedman and colleagues33 found that significant, positive relations exist between self-reported trait anxiety and subjectively measured WASO and time in bed (TIB) among older adults. On the other hand, trait anxiety was associated with increased levels of objectively measured total sleep time (TST) and TIB, indicating that trait anxiety is associated with perceptions of poor sleep quality and more desirable sleep outcomes where objective data are concerned. Similarly, McCrae et al34 found that individuals classified (according to self-reported sleep quality and sleep diaries) as “good sleepers” and “poor sleepers” without sleep complaints reported significantly lower levels of trait anxiety than those with complaints. In a related area, researchers have found that older good and poor sleepers (defined on the basis of self-report questionnaires) differed in their levels of anxiety and the relations between their self-reported life stress and sleep parameters.35 Specifically, poor sleepers reported higher levels of state anxiety than good sleepers. Interestingly, however, these groups did not differ in terms of self-reported life stress or depression. In addition, among poor sleepers, life stress was related to increased problems with sleep onset, and decreased problems with early morning awakening; these relations were not evident among good sleepers. However, among good sleepers, higher levels of state anxiety were associated with an increased number of nighttime awakenings. Fichten et al36 categorized older individuals as good sleepers, low-distress poor sleepers, and high-distress poor sleepers, on the basis of self-report. Of these three groups, good sleepers and low-distress poor sleepers reported lower levels of psychological distress on measures that included anxiety-related inventories compared to high-distress poor sleepers.

Given that DSM-IV criteria for GAD include physiological hyperarousal and insomnia, it is not surprising that older adults with GAD report elevated levels of restlessness and fatigue compared to those without anxiety symptoms.16 Interestingly, however, sleep disturbances are equally prevalent in older adults with clinical and subclinical levels of GAD symptoms.16

Taken together, research findings provide robust evidence for an association between anxiety and insomnia in the elderly. This relation appears most consistent when sleep variables are measured via self-report. Thus, when patients present with insomnia, clinicians should assess for the presence of anxiety, and vice versa. Referral to a psychologist or psychiatrist should be made when either sleep or anxiety problems cause significant distress or interfere with social or occupational functioning, and remain after treating any underlying medical problems. In addition, a sleep study should be ordered when psychiatric disturbances and chronic illnesses that commonly underlie insomnia are treated but insomnia problems remain.

PHARMACOLOGICAL MANAGEMENT OF ANXIETY: IMPACT ON SLEEP
Multiple pharmacological agents have been used to treat anxiety symptoms in older adults, and many of these medications impact sleep. The following is a brief review of studies of pharmacological management of anxiety, and the impact of these medications on sleep-related variables. Although some of the studies mentioned do not deal with anxiety per se, antidepressants and benzodiazepines are used so frequently in anxiety that it seems worthwhile to include them here.

Benzodiazepines
Sedative-hypnotics and anxiolytics are the most commonly used psychotropics in both institutional and noninstitutional settings.37 In spite of the frequency with which these drugs are prescribed, there is good reason to question the appropriateness of using these medications with older adults. Curran et al38 found that long-term use of benzodiazepines was not associated with improved sleep. These medications have been known to interfere with sleep architecture, can cause dependency, and are associated with rebound anxiety and insomnia (ie, symptoms increasing beyond pretreatment levels following discontinuation of medication). Caution is warranted when using these medications in older patients, who tend to be more sensitive to medication side effects. Furthermore, withdrawal syndromes may exacerbate pre-existing anxiety. In spite of the relatively low risk profile of short-acting benzodiazepines, they carry an elevated risk of nighttime falls in nursing home residents.39 In summary, benzodiazepines should be used intermittently, on a short-term basis, and cautiously in the elderly. For a review of the use of benzodiazepines to treat insomnia in community-dwelling elderly, the reader is referred to an article by Grad.40

Shorter-acting, nonbenzodiazepine hypnotics such as zolpidem, zaleplon, and eszopiclone may be better tolerated in this population. Intermittent use of zolpidem has shown effectiveness in chronic insomnia with benefits on nights the drug is taken; on nights without the medication, sleep is no worse than at baseline.41 Although limited information exists concerning the relative benefits of these drugs, eszopiclone appears to have a longer half-life.42 Further research is needed to support the effectiveness and safety of these medications in the elderly.

Antidepressants
In addition to treatment for depression, older antidepressants have been used for treatment of anxiety, chronic pain, fibromyalgia, migraines, and to improve sleep in patients with depression. For example, nortriptyline has been reported to improve sleep continuity.43 Trazodone is commonly administered as a treatment for insomnia. Saletu-Zyhlarz et al44 investigated the effect of trazodone on depression, anxiety, and insomnia across several age groups, and found that the medication’s most prominent effect was on symptoms of insomnia. Trazodone was also associated with decreased anxiety in this study. Haria et al45 cautioned that orthostatic hypotension must be monitored when using trazodone in the elderly. Older antidepressants have a low therapeutic index, and dosages are difficult to adjust.

The newer antidepressants also show promise in the treatment of anxiety and insomnia. Mirtazapine is sedating at lower doses and may improve sleep continuity;43,46 it may cause weight gain as a side effect. A 6-week trial of sertraline was associated with improvement in both anxiety and sleep disturbances.47 Similar findings emerged from a study of nefazodone;48 clinicians should be aware of the “black box” hepatotoxicity warning issued by the FDA. Following 24 weeks of treatment with venlafaxine, older patients with GAD (in which sleep disturbances are prominent) reported greater declines on anxiety rating scales compared to those receiving placebo.49 It should be noted that the use of some antidepressants in older adults, including selective serotonin reuptake inhibitors (SSRIs), is associated with an increased rate of falls.50 Stimulating SSRIs might increase restless legs symptoms;51 clinicians should consider timing dosage accordingly.

Other medications
Atypical antipsychotic agents, including quetiapine and olanzapine, are frequently utilized for their sedating properties. These medications are not FDA-recommended for the management of insomnia. They are most effective when symptoms of agitation accompany sleep disturbance.52 When using these agents, clinicians should monitor for excess sedation, hypotension, and metabolic syndrome.

Taken together, antidepressant medications and benzodiazepines have demonstrated therapeutic effects on target symptoms of anxiety and insomnia. However, these medications pose health risks to older adults and must be prescribed judiciously. Further research is needed to evaluate the safety and utility of different agents in the elderly.

It must be noted here that nonpharmacological interventions have been found to be effective at reducing anxiety in the elderly. For example, cognitive-behavioral therapy shows promise as a treatment for late-life GAD,53,54 in which insomnia is a very common complaint.16 Cognitive-behavioral interventions for anxiety commonly include arousal relaxation exercises (eg, muscle relaxation, imagery, breathing exercises) that have been included in effective nonpharmacological treatments for insomnia in both younger and older samples.55,56 Thus, nonpharmacological interventions for anxiety may yield concomitant improvements in sleep complaints. In light of the potential for adverse reactions to medications in the elderly, nonpharmacological interventions may well be the safest approach to the treatment of these symptoms in older adults.

CONCLUSION
A substantial body of literature has documented the high prevalence of symptoms of anxiety and insomnia among older adults. There is also increasing awareness of the common overlap between these conditions among this population in clinical settings. Further, there is ample evidence that most pharmacological interventions for anxiety also impact sleep, many times negatively. There is thus a need for systematic investigations of the linkage and interrelationships of anxiety and sleep. Specifically, we need to understand the biologic underpinnings of these syndromes and the areas of overlap and how they get translated into clinical phenomena. Such understanding should lead to development of more effective therapeutic interventions.

The research reported in this article is supported by the Veterans Affairs Palo Alto Health Care System.